Draft Final Report
Light-Duty Vehicle
Driving Behavior:
Private Vehicle Instrumentation
VOLUME 1: TECH
Presented to:
Certification Division
Office of Mobile Sources
U.S. Environmental Protection Agency
2565 Plymouth Road
Ann Arbor, Michigan 48105
Prepared by:
Radian Corporation
8501 N. Mopac Blvd.
Austin, TX 78759
24 August 1992
RADIAN
-------
CORPORATION
(Mailing Address)
P.O. Box 201088
Austin, TX 78720-1088
(Shipping Address)
8501 North Mopac Blvd.
Austin, TX 78759
(512)454-4797
DCN 92-254-036-90-04
RCN 254-036-90-06
Light-Duty Vehicle Driving Behavior:
Private Vehicle Instrumentation
Draft Final Report
Prepared for:
Jim Markey
Certification Division
Office of Mobile Sources
U.S. Environmental Protection Agency
2565 Plymouth Road
Ann Arbor, Michigan 48105
Prepared by:
T.H. DeFries
S. Kishan
Radian Corporation
8501 N. Mopac Blvd
Austin, Texas 78759
24 August 1992
-------
TABLE OF CONTENTS
Page
VOLUME 1: TECHNICAL REPORT
EXECUTIVE SUMMARY 1
ACKNOWLEDGEMENTS
1.0 INTRODUCTION 1-1
1.1 The Current Federal Test Procedure 1-2
1.2 Project Description 1-3
1.3 Structure of TTiis Report 1-5
2.0 DATALOGGING EQUIPMENT 2-1
2.1 3-Parameter Datalogger 2-1
2.1.1 3-Parameter Logger Measurements 2-2
2.1.2 Quantitative Quality Assurance Objectives 2-5
2.1.3 3-Parameter Datalogging Equipment Design 2-8
2.1.4 Laboratory Function Checks 2-11
2.1.5 Test Track Speed Accuracy Checks 2-17
2.1.6 Pilot Testing 2-36
2.2 6-Parameter Data Collection 2-39
2.2.1 6-Parameter Logger Measurements 2-39
2.2.2 Quantitative Quality Assurance Objectives
for the 6-Parameter Datalogger 2-42
2.2.3 6-Parameter Datalogging Equipment Design 2-42
2.2.4 Pilot Testing 2-53
3.0 FIELD PROCEDURES 3-1
3.1 Field Administrative Procedures 3-1
3.1.1 Introductory Leaflet 3-4
3.1.2 Cover Sheet ^-5
3.1.3 Vehicle Selection 3-5
3.1.4 Replacement Vehicle Selection 3-12
3.1.5 Interactions with Drivers 3-17
3.1.6 Vehicle Condition Check 3-20
3.1.7 Final Vehicle Acceptance 3-20
3.1.8 Handling of Driver/Owner Problems 3-23
ii
-------
TABLE OF CONTENTS (Continued)
Page
3.2 Datalogger Installation Procedures 3-23
3.2.1 3-Parameter Logger Installation 3-25
3.2.2 6-Parameter Logger Installation and Checkout 3-33
3.3 Datalogger Removal Procedures 3-43
3.3.1 3-Parameter Logger Removal 3-43
3.3.2 6-Parameter Logger Removal 3-46
3.3.3 Data Handling 3-49
4.0 FIELD RESULTS 4-1
4.1 Selection of Cities 4-1
4.2 Activities at Inspection/Maintenance Facilities 4-6
4.2.1 Spokane, Washington 4-7
4.2.2 Baltimore, Maryland 4-11
4.3 Vehicle Sampling 4-14
4.3.1 Summary of Vehicles Solicited 4-14
4.3.2 3-Parameter Instrumentation 4-35
4.3.3 6-Parameter Instrumentation 4-38
4.4 Factors Affecting the Data 4-54
5.0 DATA QUALITY CHECKING AND ARCHIVING 5-1
5.1 Data Quality Checking Methodology 5-1
5.2 Data Correction Guidelines 5-11
5.2.1 Guidelines for Spokane 3-Parameter Data Editing ... 5-12
5.2.2 Guidelines for Baltimore 3-Parameter Editing 5.14
5.2.3 Guidelines for 6-Parameter Data Editing 5-17
5.3 Data Archiving Format 5-19
5.4 Results of Data Quality Checking 5-23
6.0 DATA ANALYSIS 6-1
6.1 Statistical Analysis 6-3
6.1.1 Definitions of Terms 6-4
6.1.2 Discussion of Results 6-7
6.1.3 Comparison with Current FTP 6-38
6.2 Data Biases 6-44
6.2.1 Speed and Acceleration Biases 6-44
6.2.2 Vehicle Selection Bias 6-52
6.2.3 Datalogger Presence Bias 6-59
iii
-------
TABLE OF CONTENTS (Continued)
Page
6.3 Options for the Advanced Analysis and Simulation of
Driving Patterns 6-73
6.3.1 Advanced Statistical Analysis 6-76
6.3.2 Signal Processing Analysis 6-95
6.4 Options for Formulating Driving Cycles 6-111
6.4.1 Two Possible Bases for a Certification Test
Procedure 6-112
6.4.2 Format Options for a New Test Procedure 6-115
7.0 REFERENCES 7-1
APPENDIX A: SAMPLE QUALITY CHECKING PLOTS A-l
APPENDIX B: DATA ANALYSIS PLOTS AND TABLES .... B-l
iv
-------
TABLE OF CONTENTS (Continued)
Page
VOLUME 2: SPEED AND ACCELERATION MEASURES
All Data Without Stratification 1
FTP 21
Site 32
Spokane
Baltimore
Baltimore Station 62
Baltimore-Rossville
Baltimore-Exeter
Logger Type 88
3-Parameter
6-Parameter
Observation Phase 119
First Day
Remaining Days
Moving Only 151
Time of Day 169
0100-0600
0600-0900
0900-1600
1600-1900
1900-0100
Vehicle Type 239
Sedan, Luxury, Station Wagon
Van, Pickup, Utility
Sports Car
v
-------
TABLE OF CONTENTS (Continued)
Page
VOLUME 3: TRIP MEASURES
All Data Without Stratification 1
Data for Major Trips Only 22
FTP 44
Site 60
Spokane
Baltimore
Station 105
Baltimore-Rossville
Baltimore-Exeter
Logger Type 150
3-Parameter
6-Parameter
Vehicle Type 195
Sedan/Luxury/Station Wagon
Pickup/Van/Utility
Sports Car
Observation Phase 258
First Day
Remaining Days
Vehicle Age 303
1970s
1980s
1990s
Time of Day 365
0100-0600
0600-0900
0900-1600
1600-1900
1900-0100
vi
-------
TABLE OF CONTENTS (Continued)
Page
Time of Week 470
Weekday
Weekend
Driver Age 515
<25
25-65
>65
vii
-------
TABLE OF CONTENTS (Continued)
Page
VOLUME 4: VEHICLE DRIVER MEASURES
All Data Without Stratification 1
Data for Major Trips Only 13
Site 25
Spokane
Baltimore
Station 48
Baltimore-Rossville
Baltimore-Exeter
Logger Type 71
3-Parameter
6-Parameter
Vehicle Type 94
Sedan/Luxury/Station Wagon
Pickup/Van/Utility
Sports Car
Vehicle Age 128
1970s
1980s
1990s
Driver Age 162
<25
25-65
>65
Transmission 196
Manual
Automatic
viii
-------
LIST OF FIGURES
Page
1 Acceleration Distribution Comparison for Spokane, Baltimore,
and the FTP 6
2 Number of Vehicles with 1% of Accelerations Greater Than
Indicated Accelerations 7
3 Speed Distribution Comparison for Spokane, Baltimore,
and the FTP. 9
4 Specific Power Distribution Comparison for Spokane, Baltimore,
and the FTP 10
1-1 The LA-4 Driving Cycle of the FTP 1-4
2-1 Pressure Measured by Datalogger (kPa) 2-16
2-2 Actual Absolute Pressure (kPa) 2-18
2-3 Constant Speed Performance for Magnet Speed Sensor 2-22
2-4 Constant Speed Performance for OEM Speed Sensor 2-23
2-5 Constant Speed Performance for Speedometer Cable Sensor . ., . . 2-24
2-6 Low Speed Accuracy for Magnet Speed Sensor 2-27
2-7 Low Speed Accuracy for OEM Speed Sensor 2-28
2-8 High Speed Accuracy for Magnet Speed Sensor 2-29
2-9 High Speed Accuracy for OEM Speed Sensor 2-30
2-10 High Speed Accuracy for Speedometer Cable Sensor 2-31
2-11 Transient Performance for Magnet Speed Sensor 2-33
2-12 Transient Performance for OEM Speed Sensor 2-34
2-13 Transient Performance for Speedometer Cable Sensor 2-35
ix
-------
LIST OF FIGURES (Continued)
Page
2-14 Wiring Diagram for the 6-Parameter Non-GM Datalogger 2-45
3-1 Datalogger Installation Procedure Flow Diagram 3-2
3-2 Introductory Leaflet 3-6
3-3 Cover Sheet 3-7
3-4 Passes and Waivers Issued Per Day 3-9
3-5 Inspection Time Profile for Baltimore 3-10
3-6 Vehicle Selection Sheet 3-18
3-7 Vehicle Study Participation Agreement 3-21
3-8 Instrument Removal Appointment Form 3-22
3-9 The Volunteer Vehicle Test Program 3-24
3-10 3-Parameter Logger Installation Notes 3-26
3-11 3-Parameter Logger Calibration Sheet 3-31
3-12 6-Parameter Logger Identification and Sensor Verification 3-35
3-13 6-Parameter Logger Installation Notes 3-39
3-14 6-Parameter Logger Calibration Sheet 3-40
3-15 3-Parameter Logger Removal Report 3-44
3-16 Usage Information 3*45
3-17 3-Parameter Data Downloading Sheet 3-47
3-18 6-Parameter Logger Removal Report 3-48
3-19 3-Parameter Downloaded Data Format 3-50
3-20 6-Parameter Data Downloading Sheet (GM Only) 3-52
x
-------
LIST OF FIGURES (Continued)
Page
3-21 6-Parameter Data Downloading Sheet (Non-GM) 3-53
3-22 6-Parameter Downloaded Data Format 3-54
4-1 Howe Street Spokane Site 4-9
4-2 Howe Street Facility-Spokane 4-10
4-3 Baltimore City I/M Station 4-12
4-4 Baltimore County I/M Station 4-13
5-1 Data Quality Checking Methodology 5-2
5-2 Vehicle/Engine Operating Scenarios Relevant to Datalogger
Operation 5-15
6-1 Speed/Acceleration Contours for All Observations 6-13
6-2 Speed Distribution for All Observations 6-14
6-3 Acceleration Distribution for All Observations 6-15
6-4 Specific Power Distribution for All Observations 6-17
6-5 Trip Time Distribution for All Trips 6-19
6-6 Trip Time Distribution for Major Trips Only 6-20
6-7 Soak Time Distribution for All Trips 6-21
6-8 Soak Time Distribution for Major Trips Only 6-22
6-9 Time and Idle Distribution for All Trips 6-24
6-10 Running Time Distribution for All Trips 6-25
6-11 Time and Idle Distribution for Major Trips Only 6-26
6-12 Running Time Distribution for Major Trips Only 6-27
xi
-------
LIST OF FIGURES (Continued)
Page
6-13 Trip Distance Distribution for All Trips 6-28
6-14 Trip Distance Distribution for Major Trips Only 6-29
6-15 Daily Driven Distance Distribution 6-31
6-16 Daily Operating Time Distribution 6-32
6-17 Distribution of Number of Trips per Day for All Trips 6-34
6-18 Distribution of Number of Trips per Day for Major Trips Only . . . 6-35
6-19 Distribution of Number of Stops per Hour for All Trips 6-36
6-20 Distribution of Number of Stops per Hour for Major Trips Only . . 6-37
6-21 Speed Distribution for the FTP Cycle 6-41
6-22 Acceleration Distribution for the FTP Cycle 6-42
6-23 Specific Power Distribution for the FTP Cycle 6-43
6-24 Acceleration/Speed Contours for the FTP Cycle 6-45
6-25 Speed Profile Example 6-47
6-26 Evaluation of Calculated Accelerations for the Speedometer Cable
Speed Sensor 6-49
6-27 Evaluation of Calculated Accelerations for the Magnet Speed
Sensor 6-50
6-28 Evaluation of Calculated Accelerations for the OEM Speed Sensor 6-51
6-29 Vehicle Make Distribution of Baltimore Eligible Population 6-55
6-30 Model Year Distribution of Baltimore Eligible Population 6-56
6-31 Vehicle Make Distribution of 3-Parameter Baltimore Instrumented
Vehicles 6-57
xii
-------
LIST OF FIGURES (Continued)
Pa2e
6-32 Model Year Distribution of 3-Parameter Baltimore Instrumented
Vehicles 6-58
6-33 Vehicle Make Distribution of Solicited 3-Parameter Vehicles
That Were Replaced 6-60
6-34 Model Year Distribution of Solicited 3-Parameter Vehicles
That Were Replaced 6-61
6-35 Vehicle Make Distribution of 3-Parameter Replacement Vehicles . 6-62
6-36 Model Year Distribution of 3-Parameter Replacement Vehicles . . . 6-63
6-37 Make Distribution of 6-Parameter Eligible Baltimore Population . . 6-64
6-38 Model Year Distribution of 6-Parameter Eligible Baltimore
Population 6-65
6-39 Make Distribution of 6-Parameter Baltimore Instrumented Vehicles 6-66
6-40 Model Year Distribution of 6-Parameter Baltimore Instrumented
Vehicles 6-67
6-41 Acceleration/Speed Contour for First Observation Phase 6-69
6-42 Acceleration/Speed Contour for Second Observation Phase 6-70
6-43 Trip Time Distribution for First Observation Phase 5.71
6-44 Trip Time Distribution for Second Observation Phase 6-72
6-45 Soak Time Distribution for First Observation Phase 6.74
6-46 Soak Time Distribution for Second Observation Phase 6-75
6-47 Telecommunications Signal Transmission 6-97
6-48 Example Driving Segment 6-100
6-49 Obstacle Model for Example Driving Segment 6-108
xiii
-------
LIST OF TABLES
Page
2-1 Summary of Physical Measurements for the 3-Parameter
Datalogger 2-3
2-2 QA Objectives for Precision, Accuracy, and Completeness 2-6
2-3 Speed and RPM Function Check Results 2-14
2-4 MAP Sensor Specifications 2-15
2-5 Constant Speed Performance of 3-Parameter Datalogger
vs. Fifth Wheel 2-32
2-6 Transient Performance of 3-Parameter Datalogger vs. Fifth Wheel . 2-37
2-7 Radian Employee Vehicles Used for Austin Pilot Testing 2-38
2-8 Summary of Physical Measurements for the 6-Parameter
Datalogger 2-40
2-9 6-Parameter Datalogger Measurements 2-43
2-10 List of Domestic Models Used for the 6-Parameter Installation . . . 2-47
2-11 List of Foreign Models Used for the 6-Parameter Installation 2-52
3-1 3-Parameter Vehicle Selection Times 3-13
3-2 3-Parameter Replacement Vehicle Characteristics 3-16
3*3 3-Parameter Logger Wiring Harness Connection 3-29
3-4 Oxygen Sensor Calibration Check 3-37
4-1 Attributes of Nonattainment Areas with Centralized I/M Programs 4-4
4-2 Solicitation Summary 4-15
4-3 Summary of Field Activity Rates 4-37
4-4 Vehicles Instrumented with 3-Parameter Dataloggers 4-39
xiv
-------
LIST OF TABLES (Continued)
Page
4-5 Vehicles Solicited But Not Instrumented with 3-Parameter
Dataloggers 4-43
4-6 Vehicles Instrumented with 6-Parameter Dataloggers 4-49
4-7 Vehicles Solicited But Not Instrumented with 6-Parameter
Dataloggers 4-51
4-8 Road Condition Estimates 4-55
5-1 Suspect Observation Flagging Parameters for 3-Parameter Data ... 5-5
5-2 Suspect Observation Flagging Parameters for 6-Parameter Data ... 5-6
5-3 Data Archiving Formats 5-20
5-4 Summary of Data Quality Checking 5-24
6-1 Definition of Terms 6-5
6_2 Variable Descriptions 6-8
6-3 Summary of Statistics for All Observations in the Driving Pattern
Data Set 6-11
6-4 Speed, Acceleration, and Specific Power Statistics for
the FTP Cycle 6-39
6-5 Speed/Acceleration Distribution for the FTP Cycle 6-40
6-6 Results of Bias Analysis for Vehicle Selection 6-54
6-7 Obstacle Description Parameters 6-109
xv
-------
EXECUTIVE SUMMARY
The LA-4 driving cycle, which is an integral part of the Federal Test
Procedure (FTP), has been used for many years to certify the emissions performance of
new vehicles. This driving cycle was based on data collected in Los Angeles in the
1960s. In the 1990 Clean Air Act Amendments, Congress asked the U.S. Environmental
Protection Agency to evaluate the Federal Test Procedure, including the LA-4 driving
cycle, to determine if it is representative of today's driving. The Certification Division of
the Office of Mobile Sources has taken the lead role in conducting a study to determine
whether, in fact, the LA-4 driving cycle is representative of today's driving. To answer
the representativeness question, EPA has chosen to conduct two studies: a chase car
study and a private vehicle instrumentation study. This report documents the techniques
and the preliminary results obtained from the private vehicle instrumentation study.
In February and March of 1992, 293 privately owned light-duty vehicles
were instrumented in Spokane, Washington and Baltimore, Maryland. The instrumenta-
tion remained on each vehicle for a one-week period. Vehicle speed, engine RPM, and
manifold absolute pressure data were collected on these vehicles for every second that
the engine was on. Following the implementation of strict data quality checking
procedures, a database was created containing the one-week's worth of data from 216 of
the vehicles. This database contained information on 12,073 trips and 6,940,411 seconds
of vehicle and engine operating data.
To collect the data, special datalogging equipment was required. Two
types of dataloggers were used: a 3-parameter datalogger and a 6-parameter datalogger.
The 3-parameter datalogger was custom designed, manufactured, and tested specifically
for this project. Some of the key design requirements were that the datalogger had to be
able to be installed on any type and age of light-duty vehicle; it had to be installed in the
engine compartment; it had to be installed rapidly; it had to be used repeatedly on
different vehicles; and it had to measure vehicle speed, RPM, and manifold absolute
1
-------
pressure. The intent of these requirements was for the datalogger to be placed on many
vehicles in a relatively short period of time and for the vehicles to be sampled as close to
randomly as possible. A total of 214 vehicles were instrumented with the 3-parameter
datalogger.
A 6-parameter datalogger was provided by the Motor Vehicle Manu-
facturers Association (MVMA) and the Association of International Automobile
Manufacturers, Inc. (AIAM). This datalogger, which recorded three additional para-
meters: throttle position, coolant temperature, and equivalence ratio, was to be used in
the study in a different manner. The system tied into the engine's computer, which
required restricting the candidate vehicles to 1989-1991 vehicles of the seven manu-
facturers participating in the study: General Motors, Ford, Chrysler, Nissan, Toyota,
Mazda, and Mitsubishi. Honda joined the study in Baltimore but no Honda vehicles
were instrumented. In addition, the dataloggers could only be placed on certain models
in those model years, depending on the configuration of the power train and electronics.
A total of 79 vehicles were instrumented with the 6-parameter datalogger.
Spokane, Washington and Baltimore, Maryland were chosen as the cities
where vehicles were to be instrumented, for a variety of reasons relating to the require-
ments of the private vehicle instrumentation study and the parallel chase car study, the
studies occurred in the two cities in roughly the same time frame. Additional data that
will be used to evaluate the representativeness of the LA-4 driving cycle are being
collected in Los Angeles and Atlanta.
The dataloggers were installed on private vehicles selected at state
inspection/maintenance (I/M) stations in Spokane and Baltimore. Vehicles to have 3-
parameter dataloggers installed were selected randomly, based on the time of the day.
The selection of vehicles to have 6-parameter dataloggers installed was based on a list of
makes and models that could be instrumented. As part of the project, a solicitor
determined when a vehicle passed the I/M inspection and then approached the driver of
2
-------
the vehicle to solicit his or her participation in this project. After properly answering
some eligibility questions, drivers targeted for 3-parameter installation were offered
approximately $100, and drivers targeted for a 6-parameter installation were offered
approximately $200, to participate in the project. For the 3-parameter targeted vehicles,
if the driver agreed to participate, the vehicle was instrumented. If the driver did not
agree to participate, or the installation was not performed for any of a number of other
reasons, the characteristics of the vehicle and driver were recorded and used to select a
replacement vehicle with similar characteristics. The replacement vehicle was then
instrumented instead in an attempt to keep the vehicles instrumented with 3-parameter
dataloggers representative of the population.
Both types of dataloggers were installed on the I/M premises in an inactive
inspection lane. The 3-parameter datalogger took approximately an hour to install, and
the 6-parameter datalogger took approximately 2lh hours to install. The 6-parameter
datalogger installation involved installing an oxygen sensor on the exhaust system just
behind the catalytic converter to measure equivalence ratio.
After the datalogging equipment was installed, installation crews took the
vehicles for a short calibration and calibration check drive around the block to ensure
that the dataloggers were operating correctly. The vehicles were then returned to their
owners.
After a week of data collection, the owners brought their vehicles back for
removal of the dataloggers. In most instances, this occurred at the same I/M station
where installations had been performed. The dataloggers were removed from the
vehicles and the data were downloaded onto floppy disks for data quality checking and
data analysis.
An extensive data quality checking and archiving process was used to
convert the raw data collected from the dataloggers into good data that couid be
3
-------
subsequently analyzed. Software was written to highlight suspect data points in the data
files. Then each of the suspect data points and surrounding observations were examined
by engineering personnel to determine whether the observation should be accepted,
changed, or rejected. Some portions of the data obtained from dataloggers were found
to be questionable. The vehicles for which questionable data were found were labeled
suspect and the data were set aside for a more careful examination later.
The goal of the data analysis was to get an idea of the representativeness
of the LA-4 driving cycle by comparing its statistics with the same statistics on the large
set of data collected in this study. To do this, summary statistics were calculated for the
study and for the LA-4 cycle. However, it is clear that, although a lot of effort has gone
into calculating these statistics, the analysis presented in this report should be regarded
as exploratory. An examination of the results of this data analysis will produce many
additional questions that will need to be answered by a further examination.
The driving behavior of the vehicles can be broken down into two general
features: the detailed trip behavior of the vehicle when the engine was on, and the soak
periods when the vehicles were off. Both of these features must be characterized
because the characteristics of the tailpipe emissions of today's vehicles depend on these
features. To get an idea of the trends in the data, various statistics were calculated in
three general areas: speed and acceleration, trips, and vehicles and drivers.
When the LA-4 cycle was developed, the raw data used for the LA-4
driving cycle were modified for the FTP so that the chassis dynamometers of the day
could handle the accelerations. Accelerations with absolute values greater than 3.3
mph/s were pinned at 3.3 mph/s. Thus, the current FTP has an upper limit on the
accelerations to which it subjects automobiles, but it also demands more acceleration
time at 3.3 mph/s than did the raw data.
4
-------
Figure 1 shows a distribution of the accelerations present in the FTP and
in the data from this study from Spokane and Baltimore. The plot shows the effect of
limiting the accelerations in the FTP. The acceleration distributions for Spokane and
Baltimore are similar; however, there are accelerations that significantly exceed 3.3
mph/s. The number of observations for these exceedances is approximately 2.7% of the
total number of observations in the database.
Another way to examine the number of accelerations that exceed 3.3
mph/s is to consider the number of vehicles in this study for which any portion of the
driving had accelerations greater than 3.3 mph/s. To avoid erroneous conclusions based
on a few acceleration outliers, the 99th percentile acceleration was determined for each
of the 216 vehicles. The results are shown in Figure 2.
Figure 2 shows that 201 of the 216 vehicles (93%) had at least 1% of their
accelerations greater than 3.3 mph/s. Any point on the curve can be used to make an
analogous statement. For the 1% criteria, the curve shows a transition region between 3
and 5 mph/s where almost all vehicles have more than 1% of their accelerations greater
than 3 mph/s, but very few vehicles have more than 1% of their accelerations greater
than 5 mph/s. Each vehicle logged about 32,000 seconds worth of data during the
instrumentation week; thus, 1% of observations relate to about 320 seconds or about SVz
minutes of driving each week. A case could be made that if 93% of the vehicles have at
least 1% of their accelerations above the FTP cycle accelerations, then the FTP cycle is
not representative of today's driving patterns.
It is possible that the FTP-exceeding accelerations, which this study shows
are present in the real world, could be responsible for a disproportionate fraction of
vehicle tailpipe emissions. The present study does not contain the emissions information
for vehicles as a function of driving pattern; therefore, the effect of these acceleration
exceedances cannot be estimated here.
5
-------
Figure 1. Acceleration Distribution Comparison for Spokane, Baltimore, and the FTP
50.0
©\
(n
c
o
CO
£
CD
CO
n
O
<*—
o
c
CD
O
CD
Q_
45.0-
40.0-
35.0-
30.0-
25.0- •
20.0-
15.0-
10.0-
FTP
Spokane
Baltimore
2-3 4-5 6-7
1-2 3-4 5-6 >7
Acceleration (mph/s)
-------
Figure 2. Number of Vehicles with 1 % of Accelerations Greater Than Indicated Accelerations
Number of Vehicles with 1* of Accelerations
Greater Than Indicated
0 1 2 3 4 5 6
Acceleration (mph/s)
-------
Another parameter which is important to compare is the distribution of
vehicle speeds. The FTP speed distribution is compared with those of Spokane and
Baltimore in Figure 3. The Spokane and Baltimore speed distributions have peaks at
idle, 30-35 mph, and 55-60 mph. On the other hand, the FTP speed distribution has
peaks at idle, 20-25 mph, and 50-55 mph. Baltimore shows a larger number of observa-
tions at the higher speeds than Spokane.
Specific power is another parameter that can be compared. Specific power
is the amount of power required per unit mass of vehicle to accelerate the vehicle from a
lower speed to a higher speed. It is another measure of the demands placed on the
engine. Figure 4 shows a comparison of the specific power distribution for the FTP and
for Spokane and Baltimore as measured in this study. The figure shows that both
Spokane and Baltimore drivers demanded more specific power from their vehicles than
is represented by the FTP; this is especially noticeable for specific powers greater than
100 mph2/s. This is the high acceleration region.
A variety of other statistics were calculated for the driving in Spokane and
Baltimore. An analysis of trip measures showed that the median trip time was about 6
minutes and was smoothly distributed between 1 minute and 100 minutes. The median
trip distance was about 2.5 miles, which agrees well with the median trip distance of 3.0
miles found in the 1979 General Motors diary survey.
Because the Federal Test Procedure is based on only a 12-hour soak and a
10-minute soak, information from the engine-on period (that is, the trips) provides the
most detailed data in the procedure. Thus, the simple nature of the soak periods is
sometimes overlooked; however, it is known that the emissions produced during cold
starts can be as important or more important than emissions produced during heavy
acceleration. Soak time had a two-mode distribution. Most of the soaks had lengths
between 1 and 300 minutes. The length of a second group of soaks was between 400
and 1500 minutes; these were the overnight soaks.
8
-------
Figure 3. Speed Distribution Comparison for Spokane, Baltimore, and the FTP
FTP
~
Spokane
ESS
RQltimnro
0-5 10-15 20-25 30-35 40-45 50-55 60-65 70+
Speed (mph)
-------
Hgure 4. Specific Power Distribution Comparison for Spokane, Baltimore, and the FTP
40.0
240-260
60-80
100-120
140-160
180-200
220-240
280-300
260-280 300+
Specific Power (mpfi/s)
-------
This report presents some options for advanced analysis of the data in the
database for the purpose of understanding the driving behavior of the vehicles instru-
mented in this study, and also for developing a new Federal Test Procedure if that is
desired. Two suggestions discussed are a principal component analysis/cluster analysis
option for characterizing trips, and a signal processing/neural network approach for
characterizing segments of trips. The latter option has the capability of simulating entire
trips and soak periods. While the current Federal Test Procedure is based on a single
driving cycle, there are advantages to using a different format for a new Federal Test
Procedure, if one is needed. Some of these procedural format options are presented for
consideration.
In conclusion, it seems that the current LA-4 driving cycle has the same
shape of speed distribution as that observed in Spokane and Baltimore driving; however,
the speeds represented by the Federal Test Procedure are somewhat lower than the
speeds actually driven in Spokane and Baltimore. Also, the well-known acceleration
upper limit imposed on the LA-4 during its creation is exceeded in the Baltimore and
Spokane database. While approximately 2.7% of the observations were above the upper
limit, the effect on the emissions of today's vehicles may or may not be significant. This
depends on the relationship between driving patterns and tailpipe emissions.
11
-------
ACKNOWLEDGEMENTS
In a study of this magnitude, which dealt with the public and combined
work from many disciplines, it is appropriate that the key players and organizations
involved be thanked for their contributions.
With the assistance of the Washington and Maryland state agencies and the
inspection/maintenance contractors, the recruiting of private vehicles immediately after
exhaust inspections became routine. We especially want to thank all of the emissions
inspectors who were in the lanes for their assistance in helping us solicit vehicles, the
inspection station managers, Mark English, Peggy Vincent, and Jim Bryant, who helped
us with day-to-day station logistics, and Leo Carroll of Vehicle Test and Technology, Inc.
and Jim Brandenburg of Envirotest Technologies, Inc., who were enthusiastic about the
project and helped us get historical information for specific stations to help determine
the times of the day when vehicles should be solicited.
We would like to thank the 727 vehicle drivers who patiently listened to
our explanation of the program while they were getting their cars inspected at the I/M
stations in Spokane and Baltimore. Of these, 293 owners allowed us to attach
datalogging systems to their vehicles to acquire the data presented in this report. Many
of these vehicle owners greatly appreciated being involved in a study that may help
increase understanding of some of the environmental effects of light-duty motor vehicles.
Many people were involved with the data collection process; however, the
field installation and retrieval crews deserve special recognition. Automotive Testing
Laboratories provided Fred Home and Bill Kammerer, two experienced automotive
mechanics who proved invaluable in the field for instrumentation and removing
equipment from vehicles. Kirby Hueske and Rick Baker were the on-site solicitors at
Spokane and Baltimore who had the unusual and sometimes challenging job of soliciting
vehicle owners for participating in the instrumentation program. Mark Hutson, Jimmy
Hand, David Ranum, and Barry Walker provided the bulk of the labor by installing and
removing dataloggers from the vehicles and downloading the data onto floppy disks in
their hotel rooms at night. Sandeep Kishan was the on-site task leader in both Spokane
and Baltimore.
We would like to thank the Motor Vehicle Manufacturer's Association
(MVMA) for their contributions of automotive expertise and their piggyback project that
accompanied this EPA undertaking. Because of their support, we increased the number
of vehicles instrumented from 100 to almost 300 vehicles. Special thanks are due Greg
Walker of MVMA and Tom Darlington of General Motors, who was the FTP committee
chairman.
Finally, Radian would like to thank the EPA work assignment manager,
Jim Markey, for his understanding guidance during this interesting project, which may
have far-reaching and long-term effects on the automobiles we drive.
-------
1.0 INTRODUCTION
The Clean Air Act Amendments of 1990 require the U.S. Environmental
Protection Agency (EPA) to evaluate the Federal Test Procedure (FTP) and revise it as
necessary to ensure that vehicles are tested under circumstances that reflect actual
current driving conditions. These conditions include fuel, temperature, acceleration, and
altitude. Currently, EPA is making appropriate changes in the procedure to address
concerns about the representativeness of the fuel, temperature, and altitude provisions.
However, the current driving cycle used in the FTP has not been evaluated to determine
if it is representative of vehicles in actual use.
EPA's Certification Division of the Office of Mobile Source Air Pollution
Control has been given the lead responsibility to analyze and modify, if necessary, the
FTP. The project discussed in this document has been performed to collect information
on in-use driving patterns. An evaluation of techniques which could be used to measure
driving patterns (1) recommended that driving pattern information should be collected
using two complementary techniques: the chase car technique and the instrumentation
of private vehicles. Sierra Research has collected data using the chase car technique;
Radian Corporation has collected data from instrumented private vehicles. This report
discusses the methods, data, and preliminary results of the instrumented private vehicle
study.
The purpose of this driving modes project was to collect information on the
way a representative cross-section of U.S. light-duty vehicles is actually driven in typical
operation. The project involved instrumenting private vehicles with instrumentation
packages that measured vehicle and engine parameters on a second-by-second basis
whenever the vehicle engines were running.
The most important of the parameters measured from the instrumented
vehicles was the vehicle speed. In the data analysis, vehicle speed time profiles for the
1-1
-------
sampled vehicles are compared with the current Federal Test Procedure time profile to
determine if the FTP is representative of today's driving. Engine operating data were
also collected since they can be used to characterize the operating state of the engine in
the vehicle. From all of the data together, investigators may be able to estimate the
relative exhaust emissions of vehicles in different operating regimes.
The two main objectives of this work were:
~ To collect enough quality driving pattern data by instrumenting
private vehicles to define the driving patterns of the in-use vehicle
population; and
• To analyze the instrumented private vehicle data and compare the
measured driving patterns with patterns emerging from other
measurement techniques and with the current Federal Test
Procedure.
The development of a new prototype certification test procedure is not part of this work.
1.1 The Current Federal Test Procedure
The current Federal Test Procedure is described in the Code of Federal
Regulations (2). This procedure involves several specifications for fuel, test temperature,
preconditioning of the automobile, as well as the driving cycle used on the chassis
dynamometer. The basis for the driving cycle is the so-called LA-4, which was based on
a small driving survey made in Los Angeles in the 1960s. In the 1990 Clean Air Act
Amendments, Congress has in effect asked EPA to determine whether the LA-4 is
representative of today's driving. While vehicles have gone through several changes in
the 25 years since the LA-4 was first developed, it is not obvious that the driving patterns
of people who use vehicles in their day-to-day lives has changed. One of the reasons it is
not obvious is that driving behavior is a complex thing and it can be quantified in many
different ways.
1-2
-------
To get an idea of the complexity of driving, the current LA-4 is shown in
Figure 1-1 as a plot of vehicle speed versus time. This figure shows that the cycle is
made up of many accelerations and decelerations in an apparently random sequence. It
is not clear at first glance how a large amount of vehicle driving data such as that
collected in this study can be distilled down to a single driving cycle.
The Federal Test Procedure was originally designed to be used to test
prototype vehicles for their emission control performance. However, in reality, because
many laboratories had the ability to perform FTP testing and because of the large
amount of data generated from FTP testing, the impact of the FTP has been larger than
its originally intended purpose. Thus, it is likely that any new Federal Test Procedure
developed may also be used for other purposes.
12 Project Description
The Certification Division of EPA's Office of Mobile Sources has been
responsible for evaluating the current FTP. In the summer of 1991, the Certification
Division decided to proceed with investigating real-world driving patterns using both
instrumented private vehicles and the chase car technique. Activities on the private
vehicle instrumentation project described in this document began in earnest in August of
1991. EPA planned to instrument 50 vehicles each in Spokane, Washington and
Baltimore, Maryland. Because of the automobile industry's interest in the new Federal
Test Procedure, the Motor Vehicle Manufacturers Association (MVMA) and the
Association of International Automobile Manufacturers, Inc. (AIAM) developed a data-
logging system in pilot studies conducted earlier in 1991. When it became clear that
EPA was going to instrument vehicles in the field, MVMA/AIAM offered the experience
that they had acquired in their pilot study. As a result, in a parallel effort,
MVMA/AIAM funded a concurrent project so that the number of instrumented vehicles
in the overall study could be increased from 100 to 288. Because of project constraints,
1-3
-------
Time (s)
Figure 1-1. The LA-4 Driving Cycle of the FTP
-------
90 of these vehicles were to be instrumented with the MVMA/AIAM datalogging system
and 198 vehicles were to be instrumented with an EPA datalogging system.
The EPA datalogging system, known as the 3-parameter datalogger, was
built for EPA in the early months of the project by Radian Corporation. It was designed
to be used on any vehicle so that data on vehicle speed, engine RPM, and manifold
absolute pressure (MAP) would be collected each second the engine was on. The
MVMA/AIAM datalogging system was designed to be used only on certain late model
vehicles because it had to be tied into the vehicle engine control unit (ECU). However,
the MVMA/AIAM system collected 6 parameters: vehicle speed, engine RPM, manifold
absolute pressure, coolant temperature, throttle position, and equivalence ratio. Thus, it
was felt that a combination of the 3-parameter and 6-parameter dataloggers would
provide a good view of how vehicles are driven. Equipment preparations were
completed in January of 1992 and actual vehicle installations occurred in February and
March of 1992, in Spokane and Baltimore, respectively. A total of 293 vehicles were
instrumented, and 74% of the vehicle data files met the strict data quality requirements
of the project.
Approximately one gigabyte of data was collected in the study. This report
presents the results of a first pass analysis of this large amount of data. In addition, the
report discusses advanced methods of analysis that may be useful to understanding the
driving patterns of the vehicles. While the report does not develop candidate driving
cycles for a new Federal Test Procedure, the statistics that are calculated from the
driving patterns observed in Spokane and Baltimore should be useful for evaluating
candidate cycles proposed in the future.
1.3 Structure of This Report
This report contains several sections describing the equipment, procedures,
results, and data analyses of the vehicle driving patterns. Section 2 describes the
1-5
-------
datalogging equipment used for the collection of data with the 3-parameter dataloggers
and 6-parameter dataloggers. Section 3 describes the field procedures used to select
vehicles for datalogger installations and the detailed activities performed for datalogger
installations and removals. This section can be used by the reader to understand the
procedures used to select vehicles in a representative fashion. Section 4 describes the
selection of the cities used for the study, the facilities where the instruments were
installed on the vehicles, and the success rate of solicitation and instrumentation.
Section 5 discusses the procedures and standards used to check the quality of the driving
pattern data obtained.
Section 6 contains the data analysis. In this report, only the speed
measurements are analyzed. The other five vehicle operating parameters have been very
useful for determining the quality of the speed data, and they also may be used in the
future to determine how vehicles respond to the needs of the driver. The data analysis
considers speed, acceleration, trips, vehicle, and driver. In this report, purely statistical
measures are being considered. However, the data analysis section presents suggestions
for advanced data analysis methods that may be useful for understanding driving
patterns, simulating driving patterns, and developing new driving cycles.
1-6
-------
2.0 DATALOGGING EQUIPMENT
To determine the driving behavior of vehicles on the road, equipment was
needed to record the speed of individual vehicles and the time the vehicles were turned
on and off. The method chosen for this was to place a datalogger on selected vehicles
for a one-week period. The configuration of the datalogger was considered and the
result was that two different types of dataloggers would be used in the study: a
3-parameter datalogger and a 6-parameter datalogger.
Several attributes of the dataloggers were considered when writing the
design specifications. Some of the choices included: rent versus buy versus build, under
the hood versus in the trunk, PC based versus non-PC based, all model-year vehicles
versus late model-year vehicles, OEM sensors versus aftermarket sensors, high
temperature electrical components versus normal temperature electrical components, a
small package versus a normal-sized package, field installation versus test car lab
installation, speed only measurement versus multiple parameter measurement, and
several others. The two types of dataloggers which were used in the study were chosen
because they had different advantages and disadvantages.
The sections which follow discuss the specifications, the hardware design,
the software design, vehicle applications, laboratory function tests, test track testing, and
pilot testing of the 3-parameter and 6-parameter dataloggers. The specifications for the
datalogger operation and testing are given in the Quality Assurance Project Plan (3).
2.1 3-Parameter Datalogger
The 3-parameter datalogger took more than half of the observations in the
study. This datalogger was designed to be small and able to withstand the elevated
temperatures found underneath the hoods of vehicles so that it could be installed near to
where aftermarket speed sensors would be placed on the vehicle. The objective here
2-1
-------
was to enable to installation crew to install the datalogging system as rapidly as possible
so that the owners of the vehicles could wait while their vehicle was being instrumented.
It was believed that a short installation time and a nonintrusive method of installation
would result in a higher percentage of people who were willing to participate in the
study.
The initial concepts for the datalogger involved collecting only vehicle
speed data; however, it was later realized that the inclusion of engine RPM and manifold
absolute pressure helped to determine when the engine was on and provided a better set
of data for future analysis of the demands that drivers put on the power plant of their
vehicles in normal driving. After the data was collected, it was found that the engine
RPM and manifold absolute pressure data also helped in the data quality checking
process.
A total of 55 3-parameter dataloggers were fabricated for this project. The
first 10 dataloggers were larger than the last 45 "mini" dataloggers.
2.1.1 3-Parameter Logger Measurements
Table 2-1 gives the critical and noncritical measurements in this test
program for the 3-parameter logger. To meet the primary objectives of the project, the
critical measurements were vehicle speed and date and the time of day the engine was
turned on and turned off. Together, these three measurements determined when people
took trips, when vehicles were at rest, and how people drove their vehicles during the
trips.
Vehicle speed was collected once per second whenever the engine was
running. Depending on the vehicle type, the sensor used to measure vehicle speed was
either a Hall-effect transducer attached to the speedometer cable, the OEM vehicle
speed sensor, which is used on many late-model vehicles, or magnets and an inductive
2-2
-------
Table 2-1
Summary of Physical Measurements for the 3-Parameter Datalogger
Measurement
Measurement
Classification
Measurement
Site
Measurement
Frequency
Vehicle Speed
Critical
Speedometer Cable
Vehicle Speed
Sensor
Magnetic Pickup
Once per second
when engine is
on
Date/Time of Day
for Engine Start-
up
Critical
MAP, RPM, and
speed signals plus
software logic causes
datalogger clock to
stamp memory
Once at engine
start-up
Date/Time of Day
for Engine Shut-
Down
Critical
MAP, RPM, and
speed signals plus
software logic causes
datalogger clock to
stamp memory
Once at engine
shut-down
Engine RPM
Noncritical
Ignition coil
Once per second
when engine is
on
Manifold Absolute
Pressure (MAP)
Noncritical
Intake Manifold
Once per second
when engine is
on
2-3
-------
pickup, which was attached to a rotating vehicle drive train component such as the
driveshaft.
The date and time of day the engine was started up and shut down were
also critical measurements, because they determined when a trip took place and when
the vehicle was at rest. These times were determined once at engine startup and once at
engine shutdown.
For the 3-parameter datalogging system used in Spokane, the datalogger
considered trips to be in progress when the RPM was above 300 rpm. The datalogger
stamped the date and time in memory only at engine startup and shutdown. For the
3-parameter datalogger, which needed to sense the RPM of any vehicle from the ignition
coil, these criteria worked well in most cases, but on some vehicles a good RPM signal
was lost during the one-week instrumentation period. If this occurred, the entire vehicle
data file had to be labeled suspect, since the times of trip stop and start were uncertain.
To prevent such occurrences for the 3-parameter datalogger in Baltimore, the criteria for
the start and end of datalogging were changed by modifying the 3-parameter datalogger
software to the following:
• Begin datalogging when:
Speed >0.1 m/s
OR
RPM > 0 rpm
OR
MAP < 80 kPa
2-4
-------
• End datalogging when:
Speed = 0 m/s
AND
RPM < 300 rpm
AND
MAP > 80 kPa
Using these criteria to define the beginning and end of datalogging greatly
improved the robustness of the 3-parameter datalogger and its interface with the vehicle
systems. It also provided more information on vehicle and engine operation, which
proved useful in the data quality checking process.
Engine RPM was a noncritical measurement, but it was recorded once per
second whenever the engine was running. Engine RPM was sensed by monitoring the
frequency of the pulses on the tachometer terminal of the ignition coil. By taking into
account the number of cylinders on the engine, the number of ignition coils on the
engine, and whether the engine was a piston or rotary type, the engine RPM was
calculated and recorded in memory.
Manifold absolute pressure (MAP) was a noncritical measurement that was
logged once per second when the engine was on by using a pressure transducer on the
circuit board of the datalogger connected to any intake manifold vacuum source on the
engine by vacuum tubing.
2.1.2 Quantitative Quality Assurance Objectives
Table 2-2 lists the quality assurance objectives for the precision, accuracy,
and completeness of the parameters recorded by the datalogger. The measurements for
2-5
-------
Table 2-2
QA Objectives for Precision, Accuracy, and Completeness
Measurement
Measurement
Classification
Method
Reporting
Units
Precision*
Accuracy6
Completeness
Speed
Critical
3-parameter logger
mis
<0.16m/s
<0.16m/s
missing values per trip
RPM
Noncritical
3-parameter logger
rpm
<33 rpm
<33 rpm
£2 missing values per trip
Manifold Absolute Pres-
sure
Noncrilical
3-parameter logger
kPa
<1.6kPa + 6% of
pressure reading
<0.8kPa + 3% of
pressure reading
i2 missing values per trip
a The largest difference between two loggers being fed the same signals.
13 The largest difference between the reference signal and the datalogger recorded value.
-------
vehicle speed are based on full-scale, 12-bit resolution at 40.95 m/s. The measurements
for engine RPM and manifold absolute pressure are based on full-scale, 8-bit resolution
at 8500 rpm and 200 kPa, respectively.
The accuracy of the speed measurement depends on the resolution of the
12-bit value stored in the datalogger and on the number of pulses used to calibrate the
speed sensor on each vehicle. It is the driven distance used to calibrate the speed sensor
which limits the accuracy of the logged speed values. The calibration distance was
chosen to be around 800 feet because this produces at least 256 speed sensor pulses for
all vehicle/speed sensor configurations. Thus, the accuracy of the speed measurement is
in all cases better than (1/256) * 40.95 m/s or 0.16 m/s (= 0.36 mph).
The precision of the speed measurement is better than 0.16 m/s because of
the 12-bit resolution; however, it is probably limited by the number of pulses per second,
the times between which are averaged and inverted to get the average speed during each
second. At high speeds many time periods are averaged; at low speeds only one or two
time periods are averaged; high speeds are possibly more precise than low speeds.
Rather than be concerned with these effects on the precision of the speed measurement,
it is simplest to claim that the precision is less than 0.16 m/s.
The accuracy of the RPM assumes that errors will be present only because
of digitizing error of the least significant bit. This will be about 33 rpm. The precision
is not claimed to be better than the accuracy because the accuracy will be 1 bit. Thus, if
the accuracy criteria are met, the precision criteria will be met.
Because manifold absolute pressure will be measured in an analog fashion,
its accuracy will be no worse than the linearity of the pressure transducer [1% of full-
scale (200 kPa)] plus the error of the transducer output amplifier feedback resistors (2%
of the reading) plus the digitization error [0.4% of full-scale (200 kPa)]. Precision is
2-7
-------
dominated by the tolerances on the pressure transducer and the amplifier feedback
resistors. Thus, the precision is twice the accuracy for manifold absolute pressure.
The quality assurance objectives for the completeness of these three
parameters is that during any trip the number of missing values recorded will be less
than 2 readings. Missing values are determined by comparing the number of
observations during a trip with the difference in time stamps at the start and end of the
trip.
2.1.3 3-Parameter Datalogging Equipment Design
Radian's Electronic Services Center (RESC) designed and fabricated the
3-parameter logger. The logger contained a single circuit board to store incoming signals
on solid-state memory chips. Each logger had 768 Kb of data memory. The logger had
hardware to measure the frequency of speed transducer signals and RPM signals, an on-
board transducer to convert manifold absolute pressure to an electrical signal, and an
internal clock with its own lithium battery. The datalogger circuit board was conformal-
coated to protect it from the elements and to reduce the effects of vibration. All
components on the circuit board were 85°C-grade components. The circuit board was
provided with an RS-232 port for communicating with a laptop PC. The circuit board
was placed in a box so that the entire "mini" package measured 5 x 7 x 1.5 inches. The
inputs of the datalogger were +12 volts, ground, + speed, - speed, + ignition coil, and a
vacuum port for the engine intake manifold vacuum.
The datalogger turned on when signals from the vehicle indicated that the
engine was running. At other times, the datalogger was off and no current was drawn
from the vehicle's battery. Data were logged every second. At this logging rate for these
parameters the logger could acquire data for approximately 54 hours of engine-on time.
2-8
-------
The speed sensor used depended on how the stock vehicle was equipped.
Older cars were equipped with a speedometer cable that led from the transmission to
the back of the speedometer gauge on the dashboard. Speed sensors for most of these
vehicles were attached to the vehicle by inserting a pulse generation unit between the
transmission and the speedometer cable. These transducers were purchased from Rostra
Precision Controls; they are manufactured for after-market cruise controls. For Fords,
this type of transducer could not be inserted in the speedometer cable at the
transmission because of the method of construction of the cable. Instead, older model
year Fords required the installation of magnets on the drive shaft and the mounting of a
nearby inductive pickup to provide pulses that were sent to the datalogger. This was a
relatively easy installation; however, in a few cases it required drilling two holes in the
vehicle's sheet metal floor underneath the front seats.
Many late-model vehicles do not have a speedometer cable but are
provided with vehicle speed sensors as a way to determine the speed of the vehicle. The
speedometer gauge then converts these signals to a speed which is displayed on the
dashboard. The 3-parameter logger tied into the leads from the vehicle speed sensor
with a Scotch-Lok connector. This provided a speed signal with a pulse rate
proportional to vehicle speed.
Signals proportional to the engine RPM were obtained from the
tachometer terminal on the ignition coil. On older vehicles, coils are mounted in the
open and the terminal was accessed using a wire with a crimp connector. In many late
model cars, the coil is located inside the distributor. For this configuration, special
connectors from Thexton were used. These connectors were inserted between the
distributor electrical connection and the wiring harness. Several different types of
connectors were available for different late model vehicles.
2-9
-------
The manifold absolute pressure signal was obtained by teeing into a
vacuum line on the vehicle. Vacuum tubing was run to the transducer, which is located
on the datalogger circuit board.
Applications
The 3-parameter datalogger was used on any car selected for
instrumentation. In this study, light-duty cars and trucks with gross vehicle weights
(GVW) less than 8500 pounds were instrumented.
Data Conversion Software
Software was written to reside on the datalogger to provide several
functions:
• Enter vehicle information into memory;
• Flag the datalogger operating mode as Calibration or Run;
• Calibrate sensors on each vehicle;
• Convert signals to engineering units;
• Record data in memory;
• Conduct a parity check on each observation as it was taken;
• Display engineering units via a laptop PC as the vehicle is being
driven;
. Turn the logger on and off with the engine;
• Dump data from memory to a PC; and
• Erase memory.
2-10
-------
During each second of data collection, signals from each of the three
transducers were averaged and the average value was stored in memory. For vehicle
speed and RPM, this calculation was done by averaging the time between consecutive
pulses during each one-second period. The reciprocal of this average time between
pulses is proportional to vehicle speed or engine RPM. For manifold absolute pressure,
the average of the readings obtained during each one-second period was stored in
memory.
Software was written to provide communications between the datalogger
and a laptop PC. This allowed calibration and checkout of the datalogger after
installation on a vehicle and downloading of the data at the end of the instrumentation
period. The downloaded data format was compatible with the SAS files that were
created. The date and time were stamped only at the beginning and end of each trip to
save memory space. Between these two stamps, vehicle speed, engine RPM, and MAP
were recorded every second. SI units were used for speed, RPM, and MAP, that is, m/s,
rpm, and kPa.
A parameter was created by the datalogger software to record for each
logged observation whether the datalogger status was in the calibration mode or the run
mode. This flag was set by the installation and removal crews to differentiate between
driving by the crew and driving by the regular vehicle driver. A parameter for parity
check was also created to increase the reliability of each observation in the logged data.
The set of parameters for each observation were parity-checked and the result logged in
memory.
2.1.4 Laboratory Function Checks
The circuit design and software of the datalogger was tested to ensure that
the datalogger met the quantitative quality assurance objectives. This was done by
2-11
-------
checking a prototype datalogger in the laboratory and on vehicles. Accuracy and
precision checks were performed in the laboratory.
Following the fabrication of the 55 dataloggers, the functionality of each
unit was checked in the laboratory. These checks were in addition to the mechanical
and visual checks to which each datalogger was subjected during manufacture.
For the functional checks, software and hardware were built to provide a
rapid checkout of all datalogger functions. This ensured the proper operation of each
datalogger in the field. The functional checks included go/no go tests on each input. A
memory check program verified memory operation by recording and retrieving data.
2.1.4.1 Speed and RPM Function Tests
The accuracy of the datalogger design for speed and RPM was measured
on the test bench by feeding the prototype datalogger a signal from a square wave
generator. Simultaneously, the signal was also fed to a frequency counter wired in
parallel with the datalogger. Accuracy measurements were made using signals of similar
frequency to those found on a vehicle with an 8 cylinder, piston engine with one ignition
coil for which 800 distance pulses were obtained in 800 feet. The readings from the
frequency counter were converted to speed or RPM, as appropriate, using the following
relationships:
Speed (m/s) = [Frequency (pulses/sec)] * [0.3048 m/ft] *
[Measured Road Distance (ftY]
[No. of Pulses in Measured Road Distance (pulses)]
RPM (rpm) = 120 * [Freqi^nrv (pulses/sec)! * fNo. of Coilsl
[No. of Cylinders]
2-12
-------
The values obtained from these expressions and the values stored in the datalogger's
memory are shown in Table 2-3 and are within the quality assurance accuracy objectives
for speed and RPM, which were 0.16 m/s and 33 rpm, respectively.
2.1.4.2 Manifold Absolute Pressure Function Tests
The 3-parameter datalogger measured manifold absolute pressure using a
strain gauge type transducer, which was attached to the surface of the datalogger circuit
board. A vacuum hose from the engine was connected to the port on this transducer.
The datalogger had appropriate electronics and software to convert the signals from the
transducer to manifold absolute pressure readings. Table 2-4 shows the type of
transducer (and its specifications) used in the 3-parameter datalogger. The operating
range of the transducer was chosen so that manifold pressures of turbo-charged engines
could also be measured. The response time of the transducer was chosen so that it was
shorter than the 1-second averaging time used in the datalogger software. The following
procedures were used to check the precision and accuracy of the MAP transducer before
field activities began.
The datalogger software was written to convert the signals from the MAP
transducer to logged values as if the transducer produced accurate and precise values.
Nine dataloggers with this software and the associated hardware were manufactured.
These nine dataloggers were then checked against a two-leg mercury manometer and a
vacuum generating system to calibrate the dataloggers. Since the MAP transducers had
a useable range of 0 to 200 kPa, they were tested over as wide a range as the vacuum
and pumping system was able to produce-from 7 to 183 kPa.
The true pressure at the MAP transducer measurement port with the test
system connected was determined by adding the vertical difference between the two legs
of the mercury manometer to the local atmospheric pressure and converting to
kilopascals. Figure 2-1 shows the comparison of the actual pressure and the pressure
2-13
-------
Table 2-3
Speed and RPM Function Check Results
Generator
Logged
Expected
Input Frequency
Value in Memory
Value in Memory
(Hz)
(m/s)
(rpm)
(m/s)
(rpm)
1.48
0.45
0.45
2.92
0.89
0.89
7.35
2.24
2.24
14.66
4.46
4.47
29.37
8.94
8.95
73.5
22.39
22.40
117.48
35.79
35.81
20.00
300
300
40.02
600
600
80.01
1167
1200
160.02
2367
2400
320.4
4800
4806
400.8
6000
6012
2-14
-------
Table 2-4
MAP Sensor Specifications
Type:
SenSym SCX30AN
Precision Compensated
Pressure Sensor
Operating
Pressure Range:
0-30 psi (absolute)
Combined Linearity
& Hysteresis:
±0.1% of full scale
Long-Term Stability
of Offset & Span:
± 1% of full scale
Response Time:
100 ms
Repeatability:
±0.2% of full scale
2-15
-------
Datalogger MAP Calibration Curve
190-j
180-
o
CL
170-
160-
o>
w_
150-
140-
3
CO
130-
V)
120-
-------
logged by the datalogger. Also shown on the plot is the one-to-one line along which all
data points would be expected to fall. The results for all nine datalogger transducers are
shown in the figure, which shows that the datalogger underestimated the actual pressure
at its measurement port and that the response of the transducer had a definite curvature.
To correct for these observed effects in the transducer response, a
regression was made of the actual pressure against the datalogger pressure. A quadratic
form of the correction equation was used so that a correction could be made for the
curvature observed.
The results of this regression are shown in Figure 2-2. This plot shows the
residual of the regression (that is, the datalogger MAP minus the fitted MAP) versus the
actual absolute pressure. Each series of lines on the plot represents the response of one
transducer. It can be seen that the fit of the data forms a fan-shaped set of residual
points that increase in disparity as the actual absolute pressure increases. The lines
drawn on the plot show the ± (0.8 kPa + 3% of reading) accuracy requirements stated
in the Quality Assurance Project Plan (3). The figure indicates that these requirements
were met. The precision between any two dataloggers would be at worst the difference
between the upper data and the lower data in the figure. Thus, the precision
requirements for MAP were met.
The regression equation derived from these calibration tests was then
inserted into the datalogger software. Thus, the data values logged during data
collection were automatically corrected for the transducer calibration, and the MAP data
in the database therefore needs no correction.
2.1.5 Test Track Speed Accuracy Checks
The speed accuracy of the logged values in the 3-parameter datalogger was
tested in a vehicle on the road using a fifth wheel. Separate tests were performed using
2-17
-------
Residual Plot of Pressure Transducer Fit
Figure 2-2. Actual Absolute Pressure (kPa)
-------
the three speed sensor configurations available as inputs to the 3-parameter datalogger
system configuration. Data were collected from the fifth wheel and from the datalogger
simultaneously. Comparison of the simultaneous values indicated the accuracy of the
3-parameter datalogger as it was obtained on a vehicle.
A 1979 Buick LeSabre was equipped with a datalogger and a driveshaft
magnet speed sensor. This vehicle was also equipped with another 3-parameter
datalogger and a speedometer cable speed sensor. A 1991 Ford Taurus was equipped
with a third 3-parameter datalogger. The speed signal on this vehicle was provided by
the vehicle's OEM speed sensor wire. Each of the three dataloggers were calibrated
using standard techniques in Austin before driving to the test track.
The three vehicles were driven to the Texas Transportation Institute at
College Station, Texas, where a fifth wheel was rented for use on the Institute's test
track. The fifth wheel was manufactured by Laboratory Equipment Corporation (156
East Harrison Street, Mooresville, Indiana, 46158). The fifth wheel used the
conventional configuration with a bicycle-type wheel attached to the rear bumper of the
vehicle. The fifth wheel had a pulse generation unit attached to the axle of the wheel.
Pulses were transmitted to the electronics unit placed inside the vehicle. The electronics
unit provided a readout of speed in miles per hour to the nearest 0.1 mph. The speed
was updated about every 1/7 second.
The test track had a measured mile for the purposes of calibrating the fifth
wheel. The fifth wheel was calibrated twice before all tests and once after all tests. The
fifth wheel indicated that the measured mile had a length of 5325, 5327, and 5324 feet.
In the preparation of the speed data before analysis, the indicated speed data from the
digital readout of the fifth wheel were corrected for this slight error (0.85%) in the
measured mile.
2-19
-------
A video camera was used to record the digital display of the fifth wheel
during the test drives. The video camera had an on-screen clock which was used to
record the camera time during all test runs. This video clock was synchronized with the
datalogger clock to within approximately two seconds. This provided a means of
synchronizing the fifth wheel speeds with the datalogger speeds in an approximate
manner. Adjustment for more precise synchronizing of the data was performed in the
data analysis phase by sliding the relative time scales back and forth to achieve a best fit
of the fifth wheel and datalogger data.
Three specific drive sequences were used to test the datalogger
performance for each of the three speed sensor configurations. In general, the tests were
to determine the accuracy of the datalogger at constant speeds, both slow and fast, and
to determine the speed accuracy of the datalogger during rapid changes in speeds
encountered during accelerations and decelerations. Three types of drives were used.
Each drive was made up of a series of operating conditions. For Drives 1 and 2, each
operating condition lasted 30 seconds. For Drive 3, each operating condition lasted 15
seconds. The following describes the operating conditions sequentially for each of the
three types of drives:
Drive 1:
Turn engine on and idle.
Accelerate wide-open throttle to 50 mph and drive at a constant 50 mph.
Decelerate rapidly to 20 mph and drive at a constant 20 mph.
Decelerate rapidly to 10 mph and drive at a constant 10 mph.
Decelerate rapidly to 5 mph and drive at a constant 5 mph.
Decelerate rapidly to 2 mph and drive at a constant 2 mph.
Decelerate rapidly to 1 mph and drive at a constant 1 mph.
Stop and idle.
Turn engine off.
2-20
-------
Drive 2:
Turn engine on and idle.
Accelerate rapidly to 1 mph and drive at a constant 1 mph.
Accelerate rapidly to 2 mph and drive at a constant 2 mph.
Accelerate rapidly to 5 mph and drive at a constant 5 mph.
Accelerate rapidly to 10 mph and drive at a constant 10 mph.
Accelerate rapidly to 20 mph and drive at a constant 20 mph.
Accelerate rapidly to 50 mph and drive at a constant 50 mph.
Decelerate rapidly to 0 mph and idle.
Turn engine off.
Drive 3:
Turn the engine on and idle.
Accelerate wide-open throttle to 50 mph and drive at a constant 50 mph.
Decelerate rapidly to 0 mph and idle.
Turn engine off.
The data from the 3-parameter dataloggers were downloaded to floppy
disks in the conventional manner. The speed data from the fifth wheel for each of the
three speed sensor configurations were recovered from the videotape by examining each
frame of the video for Drives 2 and 3. The values from the videotape were transcribed
to a database for data analysis using SAS.
A comparison of the speed logging performance of the 3-parameter
datalogger and the fifth wheel can be made by examining plots and statistics. Figures
2-3, 2-4, and 2-5 show the results of Drive 2, where constant speeds of 1, 2, 5, 10, 20, and
50 mph were driven for approximately 30 seconds. The figures show, with a solid line,
the values logged by the 3-parameter datalogger and with dots, the individual data points
obtained from the fifth wheel. Only those data points for the fifth wheel for 15 seconds
during relatively constant speed are shown. The figures show very good agreement for
the magnet sensor and for the OEM speed sensor. In fact, the dots are not clearly
visible because they are on top of the solid line. They are evident as a slight broadening
2-21
-------
N>
i
NJ
K)
60.00-1
50.00
-C 40.00
Q.
E
w 30.00
~o
0)
0)
Q- 20.00
10.00-
0.00-
r
t i r i » i i i t I i 1 i1 i *T ¦> » f » *
9:56:40.000
r
9:58:20.000 10:00:00.000
Time
I I i r t ¦ i
T
10:01:40.000
Figure 2-3. Constant Speed Performance for Magnet Speed Sensor
-------
K>
N>
U)
60.00
50.00-
40.00-
CL
E
w 30.00-!
~u
Q>
-------
NJ
i
N>
-C>.
60.00 4
50.00
xT 40.00
Q_
E
w 30.00
"O
0)
CD
Q- 20.00
10.00
0.00
9:30:00.000
i v v
9:31:40.000 9:33:20.000
Time
9:35:00.000
Figure 2-5. Constant Speed Performance for Speedometer Cable Sensor
-------
of the lines at the constant speed plateaus. However, for the speedometer cable sensor,
Figure 2-5 shows that the values logged in the 3-parameter datalogger were very erratic
for speeds below 10 mph.
The speedometer cable speed sensor was attached to the speedometer
cable under the hood as it entered the cruise control transducer. Thus, there was a
length of approximately 5 feet of speedometer cable between the transmission and the
speed sensor. The erratic behavior of the data was the result of a sharp bend in the
speedometer cable which produced a jerky rotation of the inner speedometer cable as
the vehicle was driven at low speeds. Unfortunately, the results of these tests were not
known until the data analysis portion of these speed accuracy checks, and the check
could not be repeated. Since the speedometer cable speed sensor operates by the same
Hall effect principle as OEM speed sensors, the accuracy of the speedometer cable
sensor at low speeds without jerk will be as good as that of OEM speed sensors.
In the field, every effort was made to install speedometer cable transducers
in a manner which did not produce jerky motion of the inner cable. Wherever the speed
transducer was placed directly on the transmission, there should be no low speed
jerkiness in the logged values, since there is no cable between the transmission and the
speed sensor. However, in those cases where the speed sensor was placed on a cruise
control unit several feet from the transmission, there is a possibility that the logged
values will have spikes in them.
One way to look for these spikes is to examine the jerk (the second
derivative of the speed) for vehicles which used speedometer cable adapters. This was
done in the data analysis of the Baltimore data. No evidence of this type of jerk in the
speed was seen in that portion of the data quality checking; however, there may be such
spikes in the Spokane data.
2-25
-------
Figures 2-6 and 2-7 show the accuracy of the magnet and OEM speed
sensors at the low speed of 1 mph. At this speed, the speed sensor is generating about
1.5 pulses per second. These are enlarged sections of the data taken during Drive 2. It
can be seen from these plots that the 3-parameter datalogger takes accurate
measurements at these very low speeds. While such low speeds are probably used very
little in actual driving, since to maintain them required that the driver keep his foot on
the brake, these tests indicate that the datalogging system provides accurate
measurements of them.
Accuracy at high speeds is shown in Figures 2-8, 2-9, and 2-10 at 50 mph.
These plots are also an enlargement of a portion of the data taken in Drive 2. For all
three types of speed sensors, the speeds from the datalogger tracked the true speed of
the fifth wheel very well.
Several statistics were calculated for the data taken during Drive 2 for the
three types of speed sensors. These statistics are presented in Table 2-5. For each
nominal speed for each of the three speed sensors, the table shows the average and
maximum absolute speed difference between the fifth wheel and the datalogger for
1-second average values for the 15-second period of each constant speed segment. The
table shows that, except for the low speeds on the jerking speedometer cable, the
agreement is very good. The speed accuracy specification to be met is 0.16 m/s ( = 0.36
mph). The table shows this was met.
The transient performance of the 3-parameter datalogger and its sensors is
examined by considering the plots in Figures 2-11, 2-12, and 2-13. These are the results
for Drive 3, which was the wide-open throttle acceleration followed by a panic stop. It
can be seen that the datalogger values and the fifth wheel values fall closely on top of
each other.
2-26
-------
9:57:45.000 9:57:50.000 9:57:55.000 9:58:00.000 9:58:05.000
Time
Figure 2-6. Low Speed Accuracy Tor Magnet Speed Sensor
-------
11:10:45.000 11:10:50.000 11:10:55.000 11:11:00.000 11:11:05.000
Time
Figure 2-7. Low Speed Accuracy for OEM Speed Sensor
-------
50.20
50.10
50.00
49.90
*2> 49.80
g- 49.70
O, 49.60-
T) 49.50-
<1> 49.40 ^
CL
(/) 49.30
49.20 •
49.10-
49.00
48.90-
y Mi jin/u !! ! 18
I'i Ififv !!ii!i!i!l
i If flijsa '
T
T
10:00:15.000
10:00:20.000
10:00:25.000
10:00:30.000
Time
Figure 2-8. High Speed Accuracy for Magnet Speed Sensor
-------
51.00-1
N>
¦
OJ
o
CL
E
50.00-
TD
Q>
-------
50.70
50.60
50.50 ¦
50.40 -
50.30-
50.20 -
50.10
50.00
49.90
49.80-
49.70 -I
49.60
49.50
49.40-
49.30 -
49.20 -
49.10
49.00
48.90
r
1 11,1
T
9:33:45.000
9:33:50.000
9:33:55.000
Time
9:34:00.000
9:34:05.000
Figure 2-10. High Speed Accuracy for Speedometer Cable Sensor
-------
Table 2-5
Constant Speed Performance of
3-Parameter Datalogger vs. Fifth Wheel
Sensor
Type
Drive
Nominal
Speed
(mph)
Average
| Speed Difference |
(mph)
Maximum
| Speed Difference |
(mph)
Magnets
2
1
0.03
0.07
2
0.02
0.06
5
0.02
0.03
10
0.06
0.08
20
0.07
0.10
50
0.16
0.21
OEM Sensor
2
1
0.03
0.08
2
0.02
0.05
5
0.02
0.04
10
0.07
0.10
20
0.10
0.13
50
0.13
0.20
Speedometer
Cable
2
1
1.5a
9.6a
2
1.0a
2.8a
5
0.6a
2.0a
10
0.10
0.24
20
0.05
0.11
50
0.07
0.25
a Poor performance caused by speedometer cable jerking on this vehicle (see text).
2-32
-------
K)
*
U>
60.00
50.00
SZ 40.00
CL
E
w 30.00
"O
a>
a>
20.00
10.00
o.oo
« i I i v •
i I i I
-1—1—r
< I > I f I » T t
10:02:30.000
10:02:40.000 10:02:50.000
Time
10:03:00.000
Figure 2-11. Transient Performance for Magnet Speed Sensor
-------
60.00 -J
K>
I
U>
50.00
JZ 40.00
Q_
E
^ 30.00
"O
20.00
10.00
0.00-1
I 1 I t I l f
11:15:10.000
11:15:20.000 11:15:30.000
Time
T
11:15:40.000
Figure 2-12. Transient Performance for OEM Speed Sensor
-------
N)
I
U>
Ln
sz
CL
50.00
40.00
30.00
T3
$ 20.00
CL
CO
10.00-
0.00-
* I t I I I
9:35:30.000
9:35:40.000 9:35:50.000
Time
—.—-—r
9:36:00.000
Figure 2-13. Transient Performance for Speedometer Cable Sensor
-------
During high accelerations and decelerations, accurate mectsuicuicm ui
speed is more difficult than at constant speed. Under these conditions, both the fifth
wheel and the datalogger have inputs with rapidly changing pulse rates to determine the
average speed over the integrating time period. In addition, a small difference in the
relative time scales of the fifth wheel and the datalogger can mean a significant
difference in the logged speeds. Table 2-6 shows the average absolute speed difference
between the 1-second averages of the 3-parameter datalogger and the fifth wheel for
each of the Drives 3. These values are larger than the same statistics shown in Table 2-5
for constant speed operation. Examination of the comparable 1-second speed values for
the two instruments showed that the largest speed differences were seen at speeds below
10 mph, which occurred at the beginning and end of the drive. The largest four of these
for each Drive 3 are shown in Table 2-6.
2.1.6 Pilot Testing
Vehicle tests were used to ensure the completeness of the logged values,
the proper operation of the software, and the compatibility of the datalogger circuit with
the three types of speed sensors to be used in this project. Radian made these tests
using dataloggers installed on the eight Radian employee vehicles given in Table 2-7.
On-car testing of dataloggers was used to determine the performance of
the loggers in actual use conditions. Radian employee vehicles were selected to provide
a range of model years and the type of speed sensors required. The dataloggers were
left on the vehicles for a period of one week, during which the Radian employees drove
their vehicles as they normally would. At the end of the one-week period, the
dataloggers were retrieved from the vehicles and examined to determine if they had
worked properly.
Additional pilot testing of the 3-parameter dataloggers took place on four
private vehicles in the Spokane pilot study. This on-site study (4) was conducted in
2-36
-------
Table 2-6
Transient Performance of
3-Parameter Datalogger vs. Fifth Wheel
Sensor
Type
Drive
Average
| Speed Difference |
(mph)
Largest Four
| Speed Difference |
(mph)
Magnets
3
0.42
2.0, 1.2, 0.88, 0.66
OEM Sensor
3
0.65
3.5, 3.1, 2.1, 0.97
Speedometer
Cable
3
0.83
6.3, 2.0, 1.5, 1.4
2-37
-------
Table 2-7
Radian Employee Vehicles Used for
Austin Pilot Testing
Vehicle
Type of Speed Sensor
1990 Mazda 626
1971 Ford Bronco
1989 Chevrolet S-10 Blazer
1988 Honda Accord
1979 Buick LeSabre
1985 Buick Century
1980 Chevrolet Malibu
1979 Buick LeSabre
OEM Wire
Magnets
OEM Wire
Magnets
Magnets
OEM Wire
Speedometer Cable
Speedometer Cable
2-38
-------
Spokane under EPA contract 68-C1-0079 and was part of the preparation activities for
this study.
2.2 6-Parameter Data Collection
The 6-parameter datalogger took less than half of the observations in the
study; however, the datalogger was designed to collect three more parameters than the
3-parameter datalogger: throttle position, coolant temperature, and equivalence ratio, as
well as vehicle speed, engine RPM, and engine load. The datalogger was designed
around the Campbell Scientific CR10 datalogging system and, accordingly, was too large
to be placed under the hood and could not withstand underhood temperatures.
Therefore, it was mounted in the trunk of vehicles. Since this location meant it was far
from the engine control unit connectors, wiring had to be run from the front of the
vehicle to the trunk.
The objective for this datalogging system was not to install the datalogging
system rapidly, but instead to acquire a set of data which more thoroughly described
engine and vehicle operation than the 3-parameter datalogger. Because installation of
the datalogging system took much longer than an hour, the owners of the vehicles were
not asked to wait while their vehicle was being instrumented. Instead, they were given a
rental car to drive for the remainder of the day. Because of the longer installation time
and more intrusive method of installation of the 6-parameter datalogger, it was expected
that a lower percentage of people would be willing to participate in the study.
22.1 6-Parameter Logger Measurements
The critical and noncritical measurements for the 6-parameter datalogger
are given in Table 2-8. To meet the primary objectives of this project, the critical
measurements were vehicle speed, time of day, and the dates the engine was turned on
2-39
-------
Table 2-8
Summary of Physical Measurements for the 6-Parameter Datalogger
Measurement
Measurement
Classifcation
Measurement Site
Measurement
Frequency
Vehicle Speed
Critical
ECU/Diagnostic
Connection
Once per second
when engine is on
Date
Critical
Data Logger
Memory
Once per second
when engine is on
Time
Critical
Data Logger
Memory
Once per second
when engine is on
Engine RPM
Noncritical
ECU/Diagnostic
Connection
Once per second
when engine is on
Throttle Position
Noncritical
ECU/Diagnostic
Connection
Once per second
when engine is on
Manifold Absolute
Pressure
Noncritical
ECU/Diagnostic
Connection
Once per second
when engine is on
Manifold Air Flow
Noncritical
ECU/Diagnostic
Connection
Once per second
when engine is on
Coolant
Temperature
Noncritical
ECU/Diagnostic
Connection
Once per second
when engine is on
Equivalence Ratio
Noncritical
NGK UEGO
Oxygen Sensor in
Exhaust System
Once per second
when engine is on
2-40
-------
and off. Together, these three measurements determined how people took trips, when
vehicles were at rest, and how people drove the vehicle during trips.
Vehicle speed was a critical measurement collected once per second when
the engine was running. Depending on the vehicle manufacturer, the speed of the
vehicle was determined from the ECU of the vehicle or a speed sensor.
The date and time of the readings were also critical measurements.
Determination of these was based on the time stored in the datalogger. For the GM
dataloggers, this was Eastern Standard Time, and for the non-GM dataloggers it was
local time for the two cities.
All other measurements were noncritical. Engine RPM was recorded once
per second whenever the engine was running. The engine RPM was sensed by
monitoring the appropriate channel on the engine ECU from the manufacturer's
interface box. Engine load was determined by measuring manifold absolute pressure
(MAP) or mass air flow (MAF). The throttle position and the coolant temperature were
determined by monitoring the ECU.
Equivalence ratio was measured by an oxygen sensor installed in the
exhaust system. Equivalence ratio ((j>) is a measure of air/fuel (A/F) ratio relative to
the stoichiometric A/F ratio (the theoretical weight to weight ratio of the air/fuel
mixture required for complete combustion). The measured equivalence ratio is given by:
w , _ Stoichiometric A/F
Measured $ = —
Measured A/F
2-41
-------
2.2.2 Quantitative Quality Assurance Objectives for the 6-Parameter Datalogger
Table 2-9 lists the units used for the 6-parameter datalogger measurements.
The datalogger software and hardware were developed by the vehicle manufacturers.
The precision, accuracy, and completeness of the various parameters were designed by
the specific vehicle manufacturers and they differed from manufacturer to manufacturer.
Radian did not participate in designing any of the hardware or software for the
6-parameter data collection; therefore, the precision, accuracy, and completeness of the
various measurements cannot be documented by Radian.
2.2.3 6-Parameter Datalogging Equipment Design
The 6-parameter dataloggers can be divided into two categories:
• The General Motors (GM) datalogger; and
• The non-GM datalogger.
The GM datalogger was designed and built by the instrumentation group
of General Motors. The datalogger was connected to the vehicle diagnostic connector to
determine the values of the various parameters that needed to be recorded. 12 volts DC
power also had to be attached to the datalogger so that it could record the engine
parameters while the engine was running. The logger itself was also directly connected
to the oxygen sensor to record the values for equivalence ratio.
The non-GM dataloggers were developed by six different manufacturers:
Ford, Chrysler, Mazda, Mitsubishi, Nissan, and Toyota. Each of these dataloggers had a
manufacturer-specific interface box that was connected to the vehicle ECU or to the
diagnostic port on the vehicle. From this connection, the interface box interpreted the
signals for engine speed, engine RPM, throttle position, coolant temperature, and
2-42
-------
Table 2-9
6-Parameter Datalogger Measurements
Measurement
Measurement
Classification
Reporting Units
Speed
Critical
mph
Engine RPM
Noncritical
rpm
Throttle Position
Noncritical
mv
Manifold Absolute
Pressure
Noncritical
kPa
Mass Air Flow
Noncritical
kg/hr
Coolant Temperature
Noncritical
°C
Equivalence Ratio
Noncritical
= 1.000 for stoichiometric
2-43
-------
manifold absolute pressure or mass air flow. The oxygen sensor signal was also directly
connected the interface box. The interface box received the signals from the vehicle
sensors and interpreted them for the Campbell Scientific CR10 datalogger. The CR10
datalogger then used a manufacturer-specific software program to convert the electrical
impulses into engineering units for the various parameters. These values were stored in
the Campbell Scientific memory modules.
The equivalence ratio () was determined by using an NGK UEGO wide-
range oxygen sensor placed between the catalytic converter and the muffler on most
vehicles. Thus, the measurement of equivalence ratio includes combustion by the
catalytic converter, as well as by the engine. The output signal voltage from the oxygen
sensor control unit was cut in half by a resistor network. Thus, the voltage going into the
CR10 was 1.500 volts for an equivalence ratio of 1. The relationships used by the
6-parameter datalogger to convert CR10 input voltage to equivalence ratio were:
For lean conditions (A/F ^ 14.6 wt/wt, <; 1):
(J) = -1.6364 * V + 3.4546
For rich conditions (A/F < 14.6 wt/wt, > 1):
= -0.8839 * V + 2.3259
where V = the CR10 input voltage (volts)
The value of the equivalence ratio from these expressions was stored in the datalogger
memory. These relationships are linear approximations of the true curved relationship
of the oxygen sensor for equivalence ratio vs. voltage as given by the NGK specifications
sheet.
Figure 2-14 shows a wiring diagram for the non-GM 6-parameter
datalogger. Communications were only possible with the CR10 datalogger and the
memory modules. These were achieved by connecting an RS232 cable with an
appropriate interface device from the PC to the CR10 datalogger or memory module.
2-44
-------
i®J
tv*rti£ie
L.
| A*
LL
/5W
*dp
iUL-Qj.
,MFV
F^3<
4A
ft*
3 a/9 >4
|*i (foot
nr
^rjytlcHtX- CenlfiiVfeKS
-feMAtf: HS-tWM i
IU.H-5
<-o
d h r< « »ep< u»J< v
* r * 4btVt XT>^
t
*Wf/H^: *l3-»ioiA-rts-1 s Y
Hftv* : Ma-3tc\ *- t4s~1 FY
M3 3»Oi4
n
^®TAa/
ffU*J
< ttv ft »H 2»l * *» ««
£ fc X )u * £
lib
Sh
7ft
1
Figure 2-14. Wiring Diagram for the 6-Parameter Non-GM Datalogger
-------
Special Campbell Scientific software, called PC208, was used for these communications.
In addition, manufacturer-specific software was also required.
Applications
Tables 2-10 and 2-11 list the vehicles that could be used for the
6-parameter installation.
Data Conversion Software
Each 6-parameter datalogger system had manufacturer-specific software to
analyze the signals coming from the vehicle and convert them into engineering units so
that they could be stored on the datalogger. In each system, the engine parameters were
recorded for each second the engine was running.
For the GM system, the software received the signals from the vehicle
ECU and then converted the signals into engineering units and stored them on the
datalogger. These stored values could then be downloaded from the datalogger to a PC.
The software could also initialize the datalogger to erase the memory of a previous trip.
For the non-GM dataloggers, software could reside in two different places.
There could be specific software on the manufacturer interface box and on the CR10
datalogger. The interface box interpreted the electronic signals coming from the vehicle.
The box then converted these signals into a form understandable by the CR10
datalogger. The software on the CR1Q datalogger could then be used for the following:
To record the data in memory;
• To display engineering units by a laptop PC as the vehicle was being
driven;
2-46
-------
Table 2-10. List of Domestic Models Used
for the 6-Parameter Installation
General Motors
l£
Model
Enaine Tvne
YIN Digits
89
6000
2.8L
PFI
A W-K
89
6000
3.1L
PFI
A T-K
B9
6000
+
3.3L
3300
—A N-K—
89
88
3.8L
3800
H—C-K—
89
98
+•
3.8L
3800
39
Beretta
2.8L
PFI
L W-K*
89
Bonneville
4-
3.8L
3800
C-K —
39
Calais
2.3L
Quad 4
N D-K
89
Calais
2.3L
Quad 4
H.O.
N A-K
89
Calais
+¦
3.3L
3300
N—N-K
89
Camaro (Z28, IROC)
+•
*
5. 0L
PFI
p F-K—
89
Camaro (Z28, IROC)
+¦
*
5.7L
PFI
—-jr.—-8-K
89
Cavalier (Z24)
2.8L
PFI
J W-K
89
Celebrity
2.8L
PFI
A W-K—
89
celebrity
3.1L
PFI
A T-K
89
Century
2.8L
PFI
A W-K
89
Century
3.1L
PFI
A T-K
89
century
+
3.3L
3300
A W-K
89
cierra
2.8L
PFI
A W-K
89
Cierra
3.1L
PFI
A T-K
89
Cierra
3.3L
3300
A N-K
89
Corsica
2.8L
PFI
L W-K—
89
Corvette
+
*
5. 7L
PFI
Y—8-X
89
Cutlass supreme
2.5L
TBI
H R—X
89
Cutlass Supreme
2.8L
PFI
—-K W-K
89
Cutlass Supreme
3.1L
PFI
W T-K
89
Firebird (Trans An)
+
*
5.0L
PFI
—r—f-k
89
Firebird (Trans Am)
+
*
5.7L
PFI
F 8-K"
89
Grand An
2. 3L
Quad 4
89
Grand An
2.3L
Quad 4
H.O.
N A-K
89
Grand An
+
3.3L
3300
N N-K
89
Grand Prix
2.5L
TBI
W R-K
69
Grand Prix
2.8L
PFI
Vf W-K
89
Grand Prix
3.1L
PFI
W-—-T-K
89
LeSabre
+
3.BL
3800
H C-K
89
ParX Avenue
+
3.8L
3800
C—-C-K—
89
Regal
2.5L
TBI
W R-K
89
Regal
2.8L
PFI
W W-K
89
Regal
3.1L
PPI
W T-K' —
89
Skylark
+
3. 3L
3300
N N-K
89
Sunbird
2.8L
PFI
J—W-K
* m
Use ALDL pin E instead <
of
M, and connect diaq, line to nin B-
+ -
LV8 is stored in data
n place of MAP
2-47
-------
Table 2-10. (Continued)
II Model Encrlna Type VIN Check Digits
90 6000 3.1L PFI A-—T-L
90 6000 + 3.3L 3300 A N-L
90 88 + 3.8L 3800 H—OL—
90 98 + 3.8L 3800 —*C—O-L
90 Beretta 2.3L Quad 4 H.O. — L—-A-L
90 Beretta 3.1L PFI —L—T-L--—1
90 Bonneville + 3.8L 3800 —H——C-L¦ ¦
90 Calais 2.3L Quad 4 N—D-L-
90 Calais 2.3L Quad 4 H.O. N A-L
90 Calais + 3.3L 3300 N N-L
90 Camaro (Z28, IROC) 5.0L PFI F—F-L
90 Camaro (Z28, IROC) 5.7L PFI F 8-L
90 Cavalier (Z24) 3.1L PFI J T-L
9C Celebrity 3.1L PFI A—T-L
90 Century 3. 1L PFI A—T-L————
90 Century 3.3L 3300 —A N-L
90 Ciarra 3.1L PFI —A—-T-L
90 Ciarra * 3.3L 3300 —-A—N-L
90 Corsica 2 . 3L Quad 4 H.O. —~L———A-L--
90 Corsica 3.1L PFI L-—T-L-———
90 Corvette S.7L PFI Y —8-L
90 Cutlass supreme 2.3L Quad 4 —w—D-L
90 cutlass supreme 2.3L Quad 4 H.o. —w—a-l
90 cutlais supreme 2. 5L TBI —W—r-l
90 Cutlase Supreme 3.1L PFI -—w—t-L
90 Firebird (Trans Am) 5.0L PFI —F—F-L
90 Firebird (Trans An) 5.7L PFI —F—8-L
90 Grand Am 2.3L Quad 4 —N—D-L
90 Grand Am 2.3L Quad 4 H.O. —N—A-L
90 Grand Am + 3.3L 3300 —N—-N-L---*-—
90 Grand Prix 2.3L Quad 4 —W—D-L
90 Grand Prix 2.3L Quad 4 H.o. —w—A-Ir-
90 Grand Prix 2.5L TBI —w—R-L
90 Grand Prix 3.1L PFI ——T-L-
90 LeSabre + 3 • 8L 3800 —H———0—L—
90 Lumina 2.3L Quad 4 ——W—D-L- - -
90 Lumina 2.3L Quad 4 H.O. —W—A-L
9 0 Lumina 2.SL TBI W R-L
9 0 Lumina 3.1L PFI W T-L
90 Park Avenue + 3.8L 3800 -c-L —
90 R«gal 2.3L TBI W R-L
90 Regal 3.1L PFI W T-L
90 Skylark + 3.3L 3300 N N-L
90 sunbird 3.1L PFI —J—T-L—
+ ¦ LV8 is stored in data in place of MAP
2-48
-------
Table 2-10. (Continued)
is
Encine Tvpe
VIN Check Dicite
91
6000
3.11
PFI
A T-M
91
6000
3.3L
3300
A N-M
91
88
+
3.81.
3800
H C-M
91
Beretta
2.3 L
Quad
4
H-0.
1 A-*
91
Beretta
3.1L
PFI
91
Bonneville
+
3.8L
3800
H C-M
91
Calais
2.3L
Quad
4
N D-M
91
Calais
2.3L
Quad
4
H.O.
M A-M
91
Calais
+
3.3L
3300
N N-M
31
Camaro
3.1L
PFI
-—F—-Ml
91
camaro (Z28,
IBOC)
5.0L
PFI
—f—r-M
91
camaro (Z2B,
IR0C)
5.7L
PFI
p a-M-
91
Cavalier (Z24)
3. 1L
PFI
J T-M
91
Celebrity
3.1L
PFI
A T-M
91
Century
3.1L
PFI
A T-M
91
Century
+
3.3L
33O0
A N-M
91
cierra
3.1L
PFI
A -T-M
91
Cierra
•f
3.3L
3300
——A N-M
91
Corsica
2.3L
Quad
4
H. O.
91
Corsica
3.1L
PFI
11 T-M
91
Corvette
5.71
PFI
Y 8-M
91
Cutlass Supreme
2.31
Quad
4
91
Cutlass Supreme
2.51
TBI
w R-M
91
Cutlass Supreme
3.1L
PFI
W T-M
91
Cutlass Supreme
3.4L
DQHC
91
Firebird
3.1L
PFI
Y-—T-M
91
Firebird (Trans Am)
5.0L
PFI
p—f-m
91
Firebird (Trans An}
5.7L
PFI
-—F 8-M
9"
Grand An
2.3L
Quad
4
N D-M
91
Grand As
2.3L
Quad
4
H.O.
N A-M
91
Grand Am
+
3. 31
3300
N N-M
91
Grand Prix
2.3L
Quad
4
W D-M
91
Grand Pri*
2.31
TBI
W R-M
91
Grand Prix
3.11
PFI
—-W T-M———
91
Grand Prix
3.4L
DOBC
W—X-M
91
LeSabre
+
3.9L
3800
H C-M
91
Lamina
2.3L
Quad
4
V D-M
91
Lumina
2.5L
TBI
W R-M
91
Lumina
3.1L
PFI
W T-M
91
Lumina Z34
3.4L
DOHC
W X-M
91
Regal
2.51
TBI
W R-M —
91
Hegal
3.11
PFI
W Mf
91
Skylark
+
3.31
3300
N N-M
91
Sunbird
3.11
PFI
+ ¦
LV8 is stored
in data
ir. place
Of HAP
2-49
-------
Table 2-10. (Continued)
Chrysler
VEHICLE UNE
BODY
MODEL YEAR AVAILABILITY ENGINE FAMIUES AVAILABLE
89 90 91 SPANNING THESE MODEL YEARS
ENGINE SALES CODE CHART
SPIRIT
A
YES
YES
YES
EDF.EDM.EDT.EFA
EDF = 2.2L TBI -
ACCLAIM
A
YES
YES
YES
EDF,EDM.EDT,EFA
EDM = 2.5L TBI
NEW YORKER
C
YES
NO
NO
EFA
EDT = 2.5L TURBO 1
N.Y. LANDAU
C
YES
YES
NO
EFA.EGA
EDR = 2.2L TURBO IV
N Y. SALON
C
NO
YES
YES
EGA
EFA = 3.0 MMC
DYNASTY
C
YES
YES
YES
EDM.EFA.EGA
EGA = 3.3L MPI
DAYTONA
G
YES
YES
YES
EDR,EDM,EDT,EFA
EGH = 3.8L MPI
DAYTONA SHELBY
G
YES
YES
YES
EDR,EDT„EFA
LE BARON
H
YES
NO
NO
EDFtEDM,EDT,EDR
LE BARON
A
NO
YES
YES
EDM.EFA
LANCER
H
YES
NO
NO
EDF,EDM,EDT,EDR
LE BARON
J
YES
YES
YES
EDR.EDM.EDT.EFA
ARIES
K
YES
NO
NO
EDF.EDM
RELIANT
K
"YES
NO
NO
EDF.EDM
OMNI
L
YES
YES
NO
EDF
HORIZON
L
YES
YES
NO
EDF
SHADOW
P
YES
YES
YES
EDF,EDM,EDT
SUNDANCE
P
YES
YES
YES
EDF .EDM,EDT
FIFTH AVENUE
Y
NO
YES
YES
EGA.EGH
IMPERIAL
Y
NO
YES
YES
EGA.EGH
LIGHT DUTY TRUCK
CARAVAN
GRAND CARAVAN
VOYAGER
GRAND VOYAGER
s
YES
YES
YES
EDM,EDT,EFA.EGA
S
YES
YES
YES
EDT.EFA.EGA
s
YES
YES
YES
EDM,EDT.EFAtEGA
s
YES
YES
YES
EDT,EFA, EGA
2-50
-------
Ford/Mercury/Lincoln
Table 2-10. (Continued)
ESCORT
Year
1989
1990
1991
Engine Type
1.9L SOHC
1.9L SOHC
1.9L SOHC
MAP/MAF
MAP
MAP
MAF
ECU Location
By steering column
By steering column
In center counsel
TEMPO/TOPAZ
Year
1989
1990
1991
Engine Type
2.3L HSC
2.3L HSC
2.3L HSC
MAP/MAF
MAP
MAP
MAP
ECU Location
By steering column
By steering column
By steering column
TAURU
rS/S ABLE/CONTINENT AL
Year
1989
1990
1991
Engine Type
3.0L/3.8L*
3.0L/3.8L*
3.0L/3.8L*
MAP/MAF
MAP
MAP
MAF
ECU Location
Behind glove box
Behind glove box
Behind glove box
THUNDERBIRD/COUGAR
Year
1989
1990
1991
Engine Type
3.8L
3.8L
3.8L
MAP/MAF
MAP
MAP
MAF
ECU Location
Passengerside kick panel
Passengerside kick panel
Passengerside kick panel
CROWN VICTORIA/GRAND MAROU1S/TOWN CAR
Year
1989
1990
1991**
1992
Engine Type
5.0L
5.0L
5.0L/4.6L
4.6L
MAP/MAF
MAP
MAP
MAP/MAF
MAF
ECU Location
Driverside
Driverside
Driverside
* The Continental does not come in a 3.0L model
** The Town Car had a 4.6L engine in 1991 and was MAF.
2-51
-------
Table 2-11. List of Foreign Models Used for the 6-Parameter Installation
MITSUBISHI
90 Galant 2wd SOHC
89 Mirage 3 or 4 Dr. SOHC Engine
90 Mirage 3 or 4 Dr. SOHC Engine
89 Dodge/Plymouth Colt 3 or 4 Dr. SOHC Engine
90 Dodge/Plymouth Colt 3 or 4 Dr. SOHC Engine
89 Eagle Summit 3 or 4 Dr. SOHC Engine
90 Eagle Summit 3 or 4 Dr. SOHC Engine
NISSAN
89, 90 Sentra
89, 90, 91 Maxima
TOYOTA
89, 90, 91 Camry
MAZDA
89, 90, 91 626
89, 90, 91 Mx6
90, 91 Protege
90, 91 Miata
2-52
-------
• To turn the logger on and off with the engine;
• To dump data from the memory to the PC; and
• To initialize the datalogger by erasing memory.
For both kinds of 6-parameter dataloggers, the final format of the file had
exactly the same format. These files were in ASCII and were in a format compatible
with SAS.
2.2.4 Pilot Testing
Several activities were conducted to prepare for the installation of the
6-parameter dataloggers in Spokane and Baltimore. These activities included training
Radian and Automotive Testing Lab personnel by the vehicle manufacturers to install
the dataloggers, and a final checkout and installation practice in Austin of each of the
manufacturer's dataloggers.
In December of 1991, the Radian 6-parameter installation leader and the
key mechanic from Automotive Testing Labs for 6-parameter installation spent a week in
Michigan being instructed by individual member companies on the procedures for each
of the 6-parameter dataloggers for General Motors, Ford, Chrysler, Mitsubishi, Nissan,
Mazda, and Toyota. This training involved all aspects of the hardware and software that
needed to be understood by installation personnel in the field.
In this study, we have assumed that MVMA/AIAM member companies
verified the accuracy and precision of their datalogging systems through laboratory
function tests and test track speed accuracy checks. No activities specifically aimed at
these checks were part of this project.
2-53
-------
After the dataloggers were developed by each manufacturer, they were sent
to the Radian office in Austin, Texas. On-car functionality checks were then performed
on all the dataloggers from GM, Ford, and Chrysler. One manufacturer-specific vehicle
was recruited for each manufacturer, and the datalogger was installed on the vehicle.
The dataloggers were then checked in both idle and city driving modes. For the Chrysler
vehicle, a Radian employee vehicle was equipped with an oxygen sensor. For all the
other tests, the oxygen sensor was installed on a piece of exhaust pipe mounted on the
tailpipe. For at least one manufacturer, the datalogger was kept on the vehicle for a
one-week period and data were collected periodically from the vehicle to see how the
datalogger was functioning. These functionality checks were also used to train Radian
staff and to develop the procedures followed in the field. Several issues were identified
for each manufacturer and resolved before the on-site installations began in Spokane.
For Mazda, Toyota, Nissan, Mitsubishi, and Honda, no pilot testing was done. In most
cases, manufacturer representatives were available on site to resolve installation
problems.
2-54
-------
3.0 FIELD PROCEDURES
A large number of field procedures were followed to ensure the success of
the data collection project. Procedures were specified in the Quality Assurance Project
Plan (3) to ensure that the maximum number of good quality data sets would be
obtained, that private vehicle owners would be happy to participate in the project, that a
representative sample of the vehicle population would be obtained in each city, and that
all necessary information about the vehicles that were sampled would be documented for
the purposes of data quality checking and data analysis. The field procedures can be
broken down into three general activities: field administrative procedures, datalogger
installation procedures, and datalogger removal procedures. The field administrative
procedures deal primarily with the handling of the vehicle owners and the vehicles, while
the datalogger procedures deal primarily with the technical aspects of instrumenting
vehicles.
3.1 Field Administrative Procedures
Figure 3-1 shows a flowchart that describes the general installation
procedure. After the I/M test was performed, vehicles with GVWs of less than 8500
pounds that received passes or waivers were selected as target vehicles if they met any of
these criteria:
• The vehicle passed the I/M inspection at a predetermined selection
time for a 3-parameter datalogger installation;
• The vehicle met the characteristics of a refusal replacement vehicle
for a 3-parameter datalogger installation; or
• The vehicle met year/make/model criteria for a 6-parameter
datalogger installation.
3-1
-------
Car Gets l/M Test
Fails
-~ Don't Consider This Car
Passes or Waived
r
Does the car meet target criteria? ¦
Yes
No
Wait tor Another Car
Inspector Gives Leaflet to Driver and Introduces Solicitor
I
Introduction & Get Car Characteristics
Find Suitable Replacement Car/Driver
I
Are answers to driver screening questions OK?
Yes
I'
No
Does Driver Initially Agree to Participate?.
Yes
I'
No
Present detailed project description
Driver Accepts
I'
Driver Rejects
Driver signs agreement?
Yes
I"
Datalogger installed successfully?
^Yes
No
No
Pay driver consolation
Pay driver half of incentive
I
Give Driver Instruction Packet Figure 3-1. Datalogger Installation Procedure
^ Flow Diagram
Driver is on his way
3-2
-------
If the vehicle was selected as a target vehicle, the I/M inspector handed
the driver an information leaflet and introduced the driver to the Radian solicitor. The
information leaflet contained a basic description of the driving pattern project. The
solicitor asked some screening questions to determine if the vehicle's usage pattern was
suitable for participation in the program. If the vehicle's usage pattern was suitable, the
solicitor then asked the driver to participate in the program. The solicitor described the
program in some detail, answered any questions, and asked if the driver was interested in
participating. If the driver did not want to participate in the program, this brought the
solicitor's interview to an end, and the driver left the installation premises. If the driver
was interested in participating, the solicitor offered an incentive to the driver and asked
him or her to sign an instrumentation agreement.
The logger was installed, calibrated, and checked on the vehicle. The
name, address, and telephone number of the owner was documented to maintain contact
during the datalogging period. After the installation was complete, the solicitor gave the
owner an information packet that described the project, gave the driver the installation
crew's telephone numbers in case of problems, and set up an appointment for removing
the datalogger.
If a 3-parameter datalogger was installed, the driver waited for the
installation to be completed. If a 6-parameter datalogger was installed on the vehicle,
the driver was assigned a loaner vehicle, and an appointment was set up to exchange his
vehicle for the rental car.
Vehicles that participated in the driving patterns project and that had
loggers installed on them were assigned a unique number. Vehicles in Spokane were
assigned a number like S001, and vehicles in Baltimore were assigned a number like
B001. Each vehicle packet consisted of a booklet containing all the forms and sheets
required for the installation and removal process:
3-3
-------
• Cover Sheet;
• Vehicle Selection Sheet;
• Vehicle Study Participation Agreement;
• 3-parameter Logger Installation Notes;
• 3-parameter Logger Calibration Sheet;
• 3-parameter Logger Removal Report;
• Usage Questionnaire;
• 3-parameter Data Downloading Sheet;
• 6-parameter Logger Installation Notes;
• 6-parameter Logger Identification and Sensor Verification;
• 6-parameter Logger Software Check;
• 6-parameter Logger Removal Report;
• Usage Questionnaire;
• 6-parameter Data Downloading Sheet (non-GM); and
. 6-parameter Data Downloading Sheet (GM).
The introductory leaflet and each of the forms are described in the following sections.
3.1.1 Introductory Leaflet
If a vehicle passed the I/M inspection and satisfied the vehicle selection
criteria, the I/M inspector handed the driver an introductory leaflet at the end of the
inspection. The purpose of this leaflet was to introduce the drivers to the vehicle
instrumentation program in a general way so that the solicitor would not have to explain
all the details to them. The leaflet did not reveal that the test program was a measure
3-4
-------
of their driving behavior to avoid biasing driving patterns. This concept was understood
by all personnel at the instrument installation site, including I/M personnel.
The leaflet was attractive so that people understood that this test program
was a serious effort and that it was being conducted in a professional manner. A sample
leaflet is shown in Figure 3-2.
3.1.2 Cover Sheet
Each vehicle packet had a cover sheet to show which vehicle information
the packet contained. An example of the cover sheet is shown in Figure 3-3. To
facilitate vehicle tracking, a vehicle packet was used for every vehicle solicited, whether
or not the vehicle actually participated in the test program. In this way, at the end of the
test program the targeted vehicles could be compared with the participating vehicles, and
the history of vehicle selection and the reasons for vehicle nonparticipation could be
determined.
3.1.3 Vehicle Selection
Vehicles to be solicited for datalogger installation were selected from
vehicles that either passed or received a waiver during the I/M inspection. Random
vehicle selection was desired to help ensure that a representative cross-section of the
vehicle population and driver population was made. However, because of the
practicalities of vehicle instrumentation and the distribution of vehicle year, make, and
model in the population, modifications to straight random sampling had to be made.
Accordingly, vehicles for 3-parameter dataloggers were selected in a quasi-random
manner; vehicles for 6-parameter dataloggers were selected as they became available.
3-5
-------
CLEAN CARS I CLEAN AIR
The effort to produce clean vehicles and clean air
in our cities has come a long way. Today's vehicles
produce only 1/5 of the emissions of those just two
decades ago.
However, we still have a long way to go. Our cities'
air should be cleaner still, to enhance our quality
of life, and to protect our own health, as well as
our children's.
The U.S. Environmental Protection Agency
(EPA) has led the fight for clean air and cleaner
vehicles for the past 20 years. The EPA's Volun-
teer Vehicle Test Program (WTP) currently
under way will help us find better ways of control-
ling pollution from cars and trucks, while main-
taining or even improving vehicle performance.
And your cooperation in this program will help
in the effort to improve the quality of the air we
breathe.
Following the state inspection of your vehicle,
you may be asked by EPA representatives from
Radian Corporation to participate in the WTP.
Data-collecting instruments will be installed on
participants' vehicles. In return, participants will
receive a monetary incentive.
The installation of our test equipment is simple
(one example is shown below). All installations
are performed by experienced automotive techni-
cians from Automotive Testing Laboratories—
nationally recognized experts in installing test
equipment on private automobiles. The total in-
stallation takes about one hour.
Figure 3-2. Introductoiy Leaflet
3-6
-------
Vehicle Sample No. B
D®M©51®r
Datalogger Type: 3-parm 6-parm
Vehicle Selection Method:
Time of Day
Replacement for Vehicle Sample
No.
Polaroid
Photo
of Vehicle
Model Year:
Make:
Model:
Space for attaching
Owner Information
Figure 3-3. Cover Sheet
3-7
-------
For any given I/M station, it is well known that the number of vehicles
inspected per day increases toward the end of the month. This is shown graphically in
Figure 3-4, which shows the number of passes and waivers per day for Stations 3 and 4 in
Maryland. The spikes pointing downward every seventh day represent the number of
passes and waivers were issued on Saturdays; on Saturdays these I/M stations are only
open from 8 a.m. to noon.
Other than this small value seen on Saturdays, the number of cars
inspected in a given week is relatively constant. The number of cars inspected on the
first complete weeks of the month are about the same. The number of cars inspected on
the third week is greater, and during the last week the daily number of passes and
waivers is the highest.
To sample this vehicle population representatively, more vehicles should be
sampled during the last two weeks of the month than at the beginning of the month.
However, because this would have required additional staff on site to perform the
installations at the I/M station, this more representative sampling procedure was not
used. Instead, a procedure that targeted the same number of datalogger installations
every day was used.
Examination of the number of vehicles inspected for a given hour of the
day during weekdays showed that the highest use of the I/M lanes occurs between 10:00
a.m. and 2:00 p.m. This is demonstrated in Figure 3-5, which shows the relative position
of a car in relationship to the other cars inspected on a given day and the time the car is
inspected. The figure shows that in May of 1991 at this particular station in Maryland
50% of the cars were inspected by a little after noon. Since each line in the graph
represents a different weekday in an entire month, the figure shows that the profile of
inspection rates from day to day is repeatable. Specifically, the average inspection time
profile differs from the most extreme daily inspection time profile by about 40 minutes.
3-8
-------
900 *
Figure 3-4. Passes and Waivers Issued Per Day
3-9
-------
TT
s
T
0
T'
10
11
1 i
I | t I
13
T
14
Hour of the Day
select.sas
-rTr
15
TTT
16
17
-T_rr
IS
-r"Tr
19
t
£0
Figure 3-5. Inspection Time Profile for Baltimore
3-10
-------
To get a representative sample of the cars that had passed or had been
given a waiver, a random sampling of all cars on a given day should be made. This
could be done if it was known in advance how many cars on a given day would be
inspected and passed or waived. However, selection of these vehicles would involve
counting every car that came through the I/M lanes. An alternative to this procedure
was to use historical data on the distribution of times when cars were inspected to
determine the times of the day to select vehicles.
If the time is randomly selected, according to historical vehicle flow
through the inspection station, the vehicles would be close to being randomly selected.
However, the generation of such random times was found to repeatedly produce clumps
of times, meaning that up to three vehicles would have to be instrumented in a short
time. At other times of the day, no cars would have been selected and the installation
crew would be idle.
An alternative approach to determining selection times was more practical
and yet did provide a reasonable degree of randomness in the vehicles selected. In this
quasi-random approach, the historical inspection rate data for the month and station in
question was used to select times when the number of cars inspected between
consecutive times was constant. However, the time used to select the first car of the day
was randomly selected (based on historical inspection rates) from the first portion of cars
corresponding to the number of cars expected between consecutive selection times.
The result of this technique was a set of selection times for each day that
were relatively evenly spaced. The times were closer in the middle of the day than in
the morning or late afternoon because the inspection rate was higher in the middle of
the day. The times from day to day were "dithered" in time (shifter earlier or later) as a
group because of the random time selected for the first car. Each list of times was
randomly assigned to instrumentation days.
3-11
-------
The list of selection times derived by this method for Spokane and
Baltimore is shown in Table 3-1. The times for Spokane were based on Spokane
inspection records for February 1991. The times for each of the two Baltimore
inspection stations were based on inspection records for March 1991 for each of the
respective stations.
The result of this approach was that, while daily times were not random,
the times for the three-week instrumentation period were random. The advantage to the
instrumentation activities was that the work load was smoothed out so that more vehicles
could be instrumented by the same small crew. This also enabled the installation staff to
be ready for vehicles to be instrumented since they knew the selection times in advance.
Because of the restrictions on the models that could be instrumented with
6-parameter dataloggers, it was unlikely that a sufficient number of randomly-selected
vehicles would have fit the instrumentation requirements; therefore, the most workable
approach was to solicit 6-parameter target vehicles:
• When they appeared in the I/M lanes; and
• When a 6-parameter vehicle was not currently undergoing an
installation.
To meet the target number of 6-parameter installations, three vehicles on
Monday, Tuesday, Thursday, and Friday and four vehicles on Wednesday had to be
instrumented. Thus, in terms of random selection times, a 6-parameter vehicle was
selected in the morning, midday, and afternoon of most days.
3.1.4 Replacement Vehicle Selection
Every reasonable and pleasant effort was made to convince the driver of
the carefully selected 3-parameter vehicle that he and the vehicle should participate in
3-12
-------
Table 3-1
3-Parameter Vehicle Selection Times
Station
Date
Day
Times
Spokane
3 Feb 92
Monday
9:13
10:37
11:54
1:16
2:36
4:00
Spokane
4 Feb 92
Tuesday
9:45
11:04
12:22
1:43
3:06
4:32
Spokane
5 Feb 92
Wednesday
9:41
10:48
11:44
12:52
2:01
3:06
5:42
Spokane
6 Feb 92
Thursday
10:03
11:19
12:40
2:00
3:23
4:55
Spokane
7 Feb 92
Friday
9:30
10:52
12:08
1:31
2:52
4:17
Spokane
8 Feb 92
Saturday
9:49
11:09
12:30
Spokane
9 Feb 92
Sunday
Spokane
10 Feb 92
Monday
8:58
10:26
11:42
1:05
2:25
3:49
Spokane
11 Feb 92
Tuesday
9:45
11:03
12:22
1:43
3:05
3:31
Spokane
12 Feb 92
Wednesday
10:12
11:16
12:16
1:26
2:32
3:38
4:52
7:27
Spokane
13 Feb 92
Thursday
10:25
11:41
1:03
2:23
3:48
5:55
Spokane
14 Feb 92
Friday
9:26
10:48
12:04
1:27
2:47
4:12
Spokane
15 Feb 92
Saturday
10:01
11:18
12:40
Spokane
16 Feb 92 .
Sunday
Spokane
17 Feb 92
Monday
President's Day
Spokane
18 Feb 92
Tuesday
9:56
11:14
12:33
1:54
3:17
4:40
Spokane
19 Feb 92
Wednesday
9:50
10:56
11:51
1:03
2:10
3:14
4:27
6:01
Spokane
20 Feb 92
Thursday
10:08
11:23
12:44
2:04
3:28
5:03
Spokane
21 Feb 92
Friday
9:33
10:55
12:12
1:34
2:55
4:21
Spokane
22 Feb 92
Saturday
9:18
10:42
11:59
-------
Table 3-1 (Continued)
Station
Date
Day
Times
Baltimore
5 Mar 92
Thursday
9:26
10:53
12:11
1:37
3:04
4:48
Baltimore
6 Mar 92
Friday
8:22
10:21
11:39
1:03
2:28
4:01
Baltimore
7 Mar 92
Saturday
8:59
10:26
11:41
Baltimore
8 Mar 92
Sunday
Baltimore
9 Mar 92
Monday
8:13
10:17
11:36
12:59
2:24
3:57
Baltimore
10 Mar 92
Tuesday
8:37
10:27
11:45
1:10
2:34
4:09
Baltimore
11 Mar 92
Wednesday
9:34
10:59
12:18
1:43
3:12
4:57
Baltimore
12 Mar 92
Thursday
8:23
10:22
11:40
1:04
2:29
4:02
Baltimore
13 Mar 92
Friday
9:46
11:07
12:28
1:52
3:22
5:13
Baltimore
14 Mar 92
Saturday
9:03
10:29
11:44
Baltimore
15 Mar 92
Sunday
Baltimore
16 Mar 92
Monday
8:44
10:26
11:46
1:12
2:42
4:25
Baltimore
17 Mar 92
Tuesday
9:32
11:03
12:22
1:48
3:25
5:14
Baltimore
18 Mar 92
Wednesday
9:19
10:52
12:11
1:36
3:12
4:58
Baltimore
19 Mar 92
Thursday
8:52
10:31
11:50
1:16
2:47
4:31
Baltimore
20 Mar 92
Friday
8:13
10:08
11:29
12:53
2:23
4:03
Baltimore
21 Mar 92
Saturday
9:18
10:43
11:51
Baltimore
22 Mar 92
Sunday
Baltimore
23 Mar 92
Monday
9:24
10:56
12:15
1:41
3:17
5:04
Baltimore
24 Mar 92
Tuesday
9:48
11:15
12:37
2:03
3:43
5:34
Baltimore
25 Mar 92
Wednesday
8:37
10:21
11:42
1:08
2:38
4:20
-------
the study. However, if the targeted vehicles that refused to participate in the program
were simply dropped, the potential existed for a bias in the vehicles instrumented. To
attempt to remove such biases, a procedure was used to select replacement vehicles for
those 3-parameter vehicles that were targeted but that, for some reason, could not be in
the vehicle instrumentation study. The basic technique used was that solicited but
nonparticipating vehicles and drivers were characterized in a general fashion and that
replacement vehicles and drivers with the same characteristics were solicited as soon as
possible from those that received passes or waivers from the I/M inspectors.
Table 3-2 shows the four characteristics considered for each refusal
replacement and the levels that each characteristic could have. The characteristics were
designed to quickly categorize the vehicle and driver. That is, at a glance the solicitors
were able to determine the characteristics of the driver they were speaking to and of the
vehicle. The solicitors used their own judgement to determine the classification of driver
and vehicle that refused to participate and the classification of a replacement vehicle.
The solicitor did not ask the driver his age or determine the model year of the vehicle.
The idea here was not necessarily to get an exact replacement of a vehicle or driver that
refused, but to avoid replacing an unusual vehicle or driver with common ones, and vise
versa. For example, an old person driving an old pickup should not be replaced with a
young person driving a new sports car, since the driving characteristics of these two
classifications could be very different.
For 6-parameter target vehicles, refusals were not replaced according to
any category, since they were not randomly selected. Any other qualifying vehicle was
sought.
3-15
-------
Table 3-2
3-Parameter Replacement Vehicle Characteristics
Driver Age
<25
26-65
>65
Vehicle Age
New (1990 + )
Middle (1980-1989)
Old (1979-)
Vehicle Type
Luxury, Sedan, Station Wagon
Pickup, Van, Utility
Sports Car
Origin
Domestic
Foreign
3-16
-------
3.1.5
Interactions With Drivers
After the target vehicle was inspected by the I/M staff, the first
instrumentation person the driver met was the solicitor. The job of the solicitor was to
determine if the vehicle was eligible for participation, convince the driver that he or she
should participate in the vehicle instrumentation program, answer the driver's questions,
and offer the driver the incentive package. While doing this, the solicitor filled out the
Vehicle Selection Sheet in the data packet. This is shown in Figure 3-6.
The first thing the solicitor entered on the sheet was an estimated
classification of the driver's age, the vehicle's age, the vehicle type, the vehicle's origin,
and the vehicle's license number. This was done immediately in case the driver
immediately refused to participate in the program and drove off. In this way, the
solicitor had the information needed to choose a suitable replacement. The license
number was used to get the vehicle identification number (VIN) so that details of the
vehicle and engine could be obtained by decoding the VIN.
The solicitor then described the program to the driver to try to get his
participation. The solicitors established themselves as official EPA representatives.
They said that they were studying automotive emissions.
The solicitor then administered the screening questions. If the answers to
all screening questions were acceptable, the solicitor told the driver that his vehicle was
eligible to participate in the test program. The solicitor then briefly reviewed what
would happen. That is, the vehicle would be instrumented that day. The
instrumentation would be left on the vehicle for approximately one week. At the end of
the week, the instrumentation would be removed at the driver's home or place of work.
Alternatively, the driver could bring the vehicle back to the crew at the I/M station to
have the instrument removed. If it was necessary, the solicitor pointed out that
experienced mechanics performed the installation and removal. The solicitor explained
3-17
-------
Figure 3-6
VEHICLE SELECTION SHEET
©©fljgan©]'
Driver Age:
<25
26-65
>65
Vehicle Age:
Old (70s)
Middle (80s)
New (90s)
Vehicle Type:
Luxury
Pickup
Sports Car
Sedan
Van
Stn Wagon
Utility
Origin:
Domestic
Foreign
Solicitor's First Contact Date:
License Number:
Time:
Eligibility Questions:
Is the vehicle in acceptable condition for this study? Yes No
Are you the owner of record of this vehicle? Fleet vehicle. Yes No
Is this your normal, everyday car? Yes No
(no cream puffs, collector's cars, mechanic's specials)
Will you be using your car for any long trips or getting it Yes No
worked on during the next week?
Is your car running normally? Yes No
6-parm Yes No
only
Your vehicle is eligible for this study. Are you interested in Yes No
hearing about the project?
Present detailed program description.
GMs: VIN ~ Yes No
3-18
-------
to the driver that both parties would sign a promise that Radian would repair any
damages, as stated in the agreement. Finally, half of the incentive would be paid before
the installation was done if the driver agreed to participate in the program. At this point
in the approach to the driver, interaction became flexible. The solicitor had to listen to
the drivers attentively to be aware of their concerns and interests, but the solicitor was
not to stray from the established protocol. The solicitor told the driver that installation
of the instrumentation on the vehicle would not affect the performance or safety of the
vehicle while it was on, nor after it was removed.
At some point, but never before the screening questions were asked, the
solicitor offered the driver an incentive to participate in the program. The amount of
the offer depended on the type of datalogger that was to be installed, and for the 6-
parameter datalogger, depended on the manufacturer of the vehicle. For 3-parameter
dataloggers, the incentive that was offered was $100. In a few instances, owners of
vehicles asked for a little bit more money for their participations, and they were
accommodated. The incentive offer for the 6-parameter dataloggers was a bit more
flexible, because these 6-parameter dataloggers could only be installed on certain specific
makes and models and model years of vehicles. The initial offer for 6-parameter
dataloggers was $175 and, in most cases, this was sufficient to convince the drivers to
participate in the program. In some cases, the amount was raised to $200 to convince
people to participate and, in a few cases, amounts as high as $450 were required to get
participation. Half of the incentive was paid at datalogger installation and half was paid
at datalogger removal.
Some drivers were not able to decide themselves whether to have the
vehicle instrumented. Some wanted to call their spouses. In these instances, the
solicitor encouraged the driver to use the mobile telephone to make the communication.
3-19
-------
When the driver agreed to participate, as indicated by signing the
agreement, the solicitor showed the driver to the waiting room and drove the vehicle to
the mechanic.
3.1.6 Vehicle Condition Check
If the mechanic for any reason believed that a solicited vehicle was not in
safe operating condition, it was not to be accepted for participation in the program. No
vehicle was rejected, however, because of poor mechanical condition.
3.1.7 Final Vehicle Acceptance
Acceptance in the program was demonstrated by the owner's signature on
the vehicle study participation agreement shown in Figure 3-7. The agreement was also
signed by the solicitor as a representative of Radian. The document showed the amount
of incentive for the total package and the amount paid at installation. After the owner
signed the incentive agreement, the mechanic proceeded to install the datalogger on the
vehicle.
While the owner waited, the owner filled out the instrument removal
appointment sheet that documented specific information about the owner and the vehicle
for the purpose of maintaining communications and to set up a removal appointment.
This sheet is shown in Figure 3-8.
In the case of the 6-parameter logger installations, the drivers left their
cars at the I/M station for the installation of the datalogger. Drivers were provided with
rental cars to drive for the remainder of the day while the instrumentation was taking
place. Appropriate paperwork was filled out for the rental car.
3-20
-------
Figure 3-7
VEHICLE STUDY PARTICIPATION AGREEMENT
1- (OWNER) agrees to accept a total
payment of $ from Radian Corporation (RADIAN) (EPA's authorized
representative) in consideration for allowing the instrumentation of the OWNER'S
vehicle for an 8 day period. The first payment of $ will be made on
The final payment of $ will be made upon removal by RADIAN of the
installed equipment. The OWNER agrees to meet with RADIAN for the purposes of
instrument removal and to receive the final payment. In the event that RADIAN
mechanics cannot instrument the OWNER'S vehicle, RADIAN will pay the OWNER $20.
2. The OWNER is not liable for any damage to the instrumentation incurred
during normal vehicle operation. The OWNER will receive the final payment in full,
regardless of the condition of the instrumentation, given that the instrumentation
has not been disturbed or handled in any way. Manipulation or removal of the
instrumentation by anyone other than RADIAN mechanics will cause forfeiture of
the second payment.
3. RADIAN assumes responsibility for damage to the instrumented vehicle
incurred during the installation and removal of the instrumentation devices. In
addition, RADIAN assumes responsibility for any damage clearly caused by the
instrumentation during normal vehicle operation.
4. Any damage or vehicle operation problems detected must be reported to
RADIAN personnel within one week after instrument removal. RADIAN retains the
option to either provide prompt reimbursement to the OWNER for the fair market
value of the damages or to use RADIAN'S contractors for the repairs. Any repair to
the OWNER'S vehicle made in connection with this study shall be warranted by
RADIAN for thirty (30) calendar days following completion of the repair work.
5. The OWNER agrees to accept RADIAN's decision as final with regard to both
repair estimates and repairs. Other than RADIAN's warranty on workmanship
described above, the OWNER understands and agrees to waive his/her right to make
additional claims (of any kind or nature) against RADIAN one week after the date of
final payment or repair as described in paragraph 4.
X
OWNER'S
Signature
RADIAN Representative's Signature
OWNER'S
Printed Name
Date
Address
Home Telephone Number Work Telephone Number
3-21
-------
Figure 3-8
MY INSTRUMENT REMOVAL APPOINTMENT
Name:
S M T W T F S Date: Time:
My phone number at the appointment time:
I will meet the removal team at:
If I am unable to keep this appointment and want to schedule another
appointment, I can contact the mobile removal team at (509) 993-4340.
OWNER
Name
Street Address
City, State
Day Phone
Evening Phone L___
License Plate Number:
Appointment for Datalogger Removal:
Name:
S M T W T F S Date: Time:
Phone number at time of appointment:
Location:
Map Coordinates:
3-22
-------
Just before the driver left the I/M facility after the vehicle was
instrumented, he or she was given a driver information packet. The first sheet in the
packet, shown in Figure 3-9, described the test program and gave instructions on the care
of the datalogger and telephone numbers for reach the installation crew. The top half of
the appointment sheet in Figure 3-8 was given to the owner to help him remember the
appointment. The bottom half was stapled to the bottom half of the cover sheet.
Finally, the owner was given a copy of the agreements he had signed.
3.1.8 Handling of Driver/Owner Problems
It was anticipated that some drivers or vehicle owners would encounter
problems with having the datalogger on their vehicles and would need to talk to the
installation or removal teams. We wanted the owners of the vehicles to be as unworried
about the dataloggers being on their vehicles as possible, so that they would drive their
vehicles normally. When necessary, the mobile mechanic was dispatched to help drivers
with datalogger problems.
3.2 Datalogger Installation Procedures
While the datalogger installation procedures were outlined in the Quality
Assurance Project Plan (3), those procedures were not detailed enough to ensure reliable
and consistent installation and removal of dataloggers on vehicles. Accordingly, a
procedures manual was written for use by installation and removal personnel in the field.
This procedures manual contained detailed instructions and application guides for
installing RPM and speed sensors, as well as the datalogger itself, on the vehicles in the
field. Both 3-parameter and 6-parameter dataloggers were covered in the procedures
manual. The sections that follow highlight some of the important features of the 3-
parameter and 6-parameter datalogger installation and removal procedures.
3-23
-------
Figure 3-9
THE VOLUNTEER VEHICLE TEST PROGRAM
Congratulations, you have been selected as a participant in the Volunteer Vehicle Test
Program. This program is sponsored by the US Environmental Protection Agency and the
Motor Vehicle Manufacturers Association, and is being administered by Radian
Corporation, in cooperation with state and local authorities.
Your assistance is critical to the success of the program. The information obtained from
your vehicle will be used to develop new emissions standards for cars and trucks in the U.S.
Thus, your participation in the program today will put cleaner and better performing
vehicles on the road tomorrow.
The Program -- Radian is willing to offer you monetary compensation for allowing us to
install data collection devices on your vehicle. Our experienced mechanics will install
instruments on your vehicle, in most cases, while you wait. During instrument installation,
our mechanics will need to drive your vehicle around the block for checkout purposes. The
instruments will remain on the vehicle for 7 full days before our mechanics come and
remove them. Removal of our equipment will occur at a mutually agreed upon time and
place. At that time, our representatives will ask you a few follow-up questions and provide
you with the final payment for your cooperation.
Instrumenting Your Vehicle ~ Our monitors are passive, data-collecting devices - that is,
nothing should interfere with the normal operation of your car or truck. No changes will
be made to the engine, electrical, or exhaust systems without prior permission from you, the
vehicle owner. The presence of the devices should in no way affect the performance,
efficiency, or safety of your vehicle.
Problems with Your Vehicle — In the unlikely event that your vehicle has operations
problems while the instrument is in place, and there is reason to suspect that our equipment
or installation is at fault, our mobile mechanics can be reached at the number provided
below. If necessary, a mechanic will be dispatched to you and your vehicle immediately.
Though the equipment presents no risk to you, you should not touch, disturb or handle the
instrument in any way while it is on your vehicle. Such handling may disturb the instrument
and invalidate the data. Manipulation or removal of our equipment by anyone other than
our own mechanics may cause forfeiture of the second cash payment.
Questions - If you have any farther questions while the instrument is in place, you can call
either of the following numbers for assistance:
For instrument removal appointment changes: (509) 993-4358
If no answer; (509) 993-4340
3-24
-------
3.2.1 3-Parameter Logger Installation
The mechanic was the person in charge of installing the dataloggers on the
private vehicles. He was assisted by the mechanic's assistant, who also calibrated the
datalogger after it was installed. When the mechanic completed the installation, he
completed the 3-Parameter Logger Installation Notes sheet, which is shown in Figure
3-10. The purpose of the checklist was to inspect specific aspects of the installation and
to make a note of what was done to the vehicle so that the it could be restored to its
original condition when the datalogger was removed.
3.2.1.1 Sensor Installation
The speed sensor was chosen from among the available Rostra Precision
Controls (RPC) components to fit the model, make, and model year of the vehicle being
instrumented. In general, the speed sensors were attached to the speedometer cable at
the transmission, inductively coupled to magnets installed on the drive shaft, or tied in
electrically to the output of the OEM speed sensor. The installation mechanic
determined the type of speed sensor to be used on a vehicle by examining the vehicle
and by consulting the RPC speed sensor application guides.
All older model vehicles and some late model vehicles use speedometer
cables. Where these can be easily unscrewed at the transmission, the after-market speed
sensors were used. Speed sensors were not installed on the speedometer cable at the
back of the speedometer gauge because of the possibility of disturbing dashboard
components and because of the difficulty of reaching this location on many vehicles. The
speedometer cable installation was done in some cases by accessing the speedometer
cable under the hood when an OEM cruise control was on the vehicle. Late-model
vehicles that do not have speedometer cables are equipped with OEM speed sensors,
and the wire from the speed sensor to the ECU was tied into with a Scotch-Lok.
3-25
-------
Figure 3-10
3-PARAMETER LOGGER INSTALLATION NOTES
DISTANCE SENSOR
Type: Speedometer Cable Adapter (RPC )
Driveshaft Magnets (RPC 4165)
OEM Speed Sensor Wire
Point of Attachment:
RPM SENSOR
Type: Thexton Coil Adapter No.:
Other:
Comments:
VACUUM SOURCE
Comments:
Vehicle location of vehicle parts removed during the installation:
3-26
-------
RPM signals were sent to the datalogger from the tachometer terminal of
the coil. Commercially available adapters from Thexton were used to access this signal.
On late model cars, the Thexton adapters fit between the connector on the distributor
and the wiring harness connector. A mechanic checked the signal from the RPM
connector with a tachometer to ensure that the tachometer terminal of the coil had been
selected for transmitting the signal to the datalogger.
The mechanic selected an appropriate source of manifold vacuum to route
to the vacuum port on the datalogger. The mechanic checked the vacuum signal with a
standard vacuum gauge. An existing vacuum line was disconnected and a tee inserted to
provide a source for the datalogger. An additional tee was placed in the vacuum line
that leads from the existing line to the datalogger for the calibration check of the
pressure transducer in the datalogger against the mechanic's vacuum gauge.
3.2.1.2 Box Installation
The 3-parameter logger was installed at a location distant from heat and
moving parts that was not subject to much vibration and that could not be struck directly
by rocks, water, dirt, or oil. The datalogger was attached to the vehicle under the hood
in a suitable location with tie wraps.
3.2.1.3 Wires
Wires and vacuum lines were routed from sensors to the datalogger to
avoid areas where damage could occur from hot parts and moving parts. Leads were
securely tie-wrapped to existing parts of the vehicle so that they were not damaged.
Standard crimp connectors were used to join sensors to datalogger input wires. Color-
coded wires were used to avoid confusion about the inputs to the datalogger. The power
for the datalogger was from a wire run directly to the +12 volts on the battery. An in-
line power fuse on the +12 volt line protected the vehicle's electrical system. Datalogger
3-27
-------
ground was also connected directly to the negative terminal of the battery. The
connection points for the 3-parameter logger wiring harness are shown in Table 3-3.
3.2.1.4 Troubleshooting
If the datalogger was installed properly, it worked when powered up. If the
datalogger did not work, then normal troubleshooting procedures were performed to
make sure that all systems were operating properly. If the datalogger could not be made
to work, there was no point in letting the vehicle participate in the test program, and the
datalogger and sensors were removed.
3.2.1.5 3-Parameter Logger Calibration and Calibration Check
Because of differences in the RPM and speed sensors, and because of
individual differences between vehicles, the RPM and speed sensors were calibrated after
installation. Calibration of the datalogger was performed by the mechanic's assistant
with the mechanic or the solicitor's assistant driving the car.
The datalogger was checked for calibration and proper operation on the
vehicle by plugging an RS-232 adaptor cable into the datalogger and running the cable
inside the passenger compartment and connecting to a laptop PC. Checks were made
against the mechanic's instruments and against the vehicle's instruments as the vehicle
was driven and the engine was run.
The first thing that the calibrator did after connecting the PC to the
datalogger was to make certain the datalogger was in the calibration (CAL) status. Then
he entered the essential information about the vehicle identification. He was prompted
by software to enter sample ID, VIN, model year, make, model, transmission type, and
beginning odometer reading.
3-28
-------
Table 3-3
3-Parameter Logger Wiring Harness Connection
Color Connection Point
Red Battery + (12 volt)
Black Battery - (Ground)
Blue Speed Signal
• Speedometer Cable Adap" ir
(Blue wire)
• Magnetic Pickup (Blue wire)
• OEM Vehicle Speed Sensor Wire
Gray Speed Return Signal
• Speedometer Cable Adaptor
(Gray wire)
• Magnetic Pickup (Gray wire)
White Ignition Coil Tachometer Terminal
3-29
-------
The clock chip in the datalogger was set to Central Standard Time in the
laboratory before field work began. The clock chip has an internal battery good for 3
years at 85°C; however, the date and time setting was checked after installation of the
datalogger against the watch of the calibrator to ensure it was reasonably correct. These
data were entered with the rest of the 3-parameter logger calibration data on the 3-
Parameter Logger Calibration Sheet, shown in Figure 3-11.
The calibration of engine RPM required the mechanic's assistant to answer
questions on the PC provided by the datalogger software. The information to be
supplied was the number of cylinders in the engine, the number of coils on the engine,
and whether the engine was a conventional Otto or a Wankel engine configuration. No
diesel engines were instrumented in this study. From the answers to these questions, the
software calculated a conversion factor to be used to convert the pulses obtained from
the tachometer terminal of the ignition coil to RPM readings.
Once these questions were answered, a calibration check of RPM was
made by comparing the mechanic's hand-held tachometer with the RPM indication
shown on the laptop computer. The RPM was checked at idle and at fast idle. The
values given by the datalogger and by the mechanic's tachometer were recorded on the
calibration sheet.
In a few instances, it was found that these RPM values disagreed by a
substantial amount. In this situation, the installers checked the datalogger installation,
checked the answers to the datalogger questions about engine configuration, checked the
tachometer switch settings, and, if necessary, tried another datalogger. In some cases, an
RC filter was inserted in the RPM input wire to the datalogger. This filter caused the
waveform voltage to be shifted down. In many problem cases, this caused the RPM
input circuit to trigger properly and to produce an accurate and reliable RPM reading.
In Spokane, for installations where a good RPM signal could not be obtained, the
dataloggers were removed from the vehicles, because the start and end of trips depended
3-30
-------
Figure 3-11
3-PARAMETER LOGGER CALIBRATION SHEET
S©]Jg31©T
Datalogger Number
has been installed in the vehicle.
Put datalogger in CAL mode.
Vehicle Sample No. : S
VIN: ~
Model Year:
Make:
Model:
_ B (get from Cover Sheet)
(Fill for 1981+cars)
(get from label under the hood)
Transmission:
Odometer:
Automatic Manual: 3 4 5 6
Measurement
Vehicle's
Mechanic's
Toler-
Datalogger
Instruments
Instruments
ance
Reading
Date
Time (CST)
Vacuum at Idle
RPM at Idle
RPM at Fast Idle
Speed (> 30 mph)
" "
±1
±67
±67
±6 %
Perform WOT acceleration from <10 mph to 35-60 mph. Local Time
Explanation of calibration drive abnormalities and corrective action:
Put datalogger in RUN mode before releasing vehicle to driver.
3-31
-------
critically on a good RPM reading. On the other hand, because of the improved trip
sensing logic in Baltimore, vehicles could be safely released. There, trip starts and ends
were reliably detected with the redundance provided by the MAP and speed sensors.
The manifold vacuum was checked at idle by monitoring the datalogger via
the PC. At the same time, the mechanic obtained a manifold vacuum reading by tying
into the tee on the vacuum line leading into the datalogger. These comparable values
were recorded on the calibration sheet and should have been reasonably close to each
other. Because of errors in the mechanic's vacuum gauge and changes in atmospheric
pressure, they were not exactly the same. These MAP checks were only calibration
checks since the response of the pressure transducer on the datalogger could not be
changed. The information that a MAP sensor was reading incorrectly during this
calibration check would have been used to delete the MAP values from the data for that
vehicle. Since MAP was not a critical parameter for the study, an incorrectly reading
MAP sensor would not have been a sufficient reason for removing the datalogger from
the vehicle at the end of installation. The critical speed data would still be good.
The speed sensor was calibrated by driving the vehicle over a measured
distance of about 800 feet on pavement. During the calibration phase of the speed
sensor, the datalogger counted pulses that the speed sensor put out for the measured
distance. The calibrator told the datalogger when the datalogger should begin and stop
counting the pulses for this short drive. The speed at which the vehicle drove was not
important. At the end of the drive, the software asked the calibrator the distance the
vehicle had driven and, based on that distance, a calibration factor was calculated by and
stored in the datalogger. Then the vehicle was driven across a second measured distance
to check the calibration. The distance that the datalogger indicated the vehicle had
driven had to agree to within 1.0% of the true distance.
Finally, the vehicle was driven on the street, and the indicated speed on
the vehicle speedometer was compared with the speed indicated on the laptop. Those
3-32
-------
two speeds had to agree within 5-10 percent. The speedometer could read incorrectly
because of tire sizes, which may have been different from those originally on the vehicle.
This was only a calibration check of the speed as measured by the datalogger, since
vehicle speedometers are known to be inaccurate to a degree.
The final part of the datalogger calibration involved doing a wide-open
throttle acceleration of the vehicle. This provided a measure of the vehicle's ultimate
performance for comparison with acceleration versus speed profiles in the data analysis
phase. In addition, this provided a check of the MAP sensor since MAP should be about
100 kPa during a wide open throttle acceleration and about 30 kPa at idle. Usually, a
0-40 miles/hour acceleration was performed.
32.1.8 Memory, Data, Downloading Check
After the installation calibration drive, the data collected during the drive
were downloaded to the laptop PC and inspected to see if they were reasonable. This
was the final datalogger check before the vehicle was released to its owner. The
datalogger status flag, which differentiates calibration driving from owner driving, was
switched from CAL to RUN just before giving the vehicle to the owner.
3.22 6-Parameter Logger Installation and Checkout
The MVMA/AIAM wanted to have a certain distribution of manufacturers
represented with the 6-parameter datalogger installations. To accomplish .this, a list of
target make, model year, and models for each manufacturer was maintained. As vehicles
were instrumented, the list was updated to indicate that certain slots had been filled.
Any remaining slots that were open showed the types of vehicles that could be used to
fill the quota.
3-33
-------
In addition, since the 6-parameter dataloggers were specific to certain
manufacturers and their vehicles, a current inventory list of the vehicle-specific
datalogger systems on hand was maintained.
Seven domestic and foreign vehicle manufacturers participated in the
driving modes project. As a vehicle entered the I/M lane and was selected to be a
candidate vehicle for the 6-parameter datalogger, the vehicle characteristics were
compared with the list of vehicles that could be instrumented. If all the vehicle
characteristics matched those of a vehicle that could be instrumented with a 6-parameter
logger, the solicitor tried to persuade him to participate in the project. If the driver
agreed, the solicitor filled out the top half of the 6-Parameter Logger Identification and
Sensor Verification form shown in Figure 3-12.
The solicitor then drove the vehicle to the 6-parameter installation area.
The mechanic's assistant filled out the bottom part of Figure 3-12, in which the details of
the selection characteristics used to identify the 6-parameter vehicle were noted. Before
the instrumentation process began, the error codes on the vehicle were checked to make
sure that all vehicle sensors were operating properly. If error codes were found, the
vehicle was not instrumented. An initial determination was made that the vehicle
exhaust system was in proper working order.
For each 6-parameter vehicle, the speed, RPM, MAP, coolant temperature,
and throttle position sensors were installed first. Once these were confirmed to be
working appropriately, the vehicle system was tested for the oxygen sensor channel. This
procedure was done before going to the muffler shop and before the oxygen sensor boss
was installed on the vehicle tailpipe. The oxygen sensor signal in the power cables was
connected to the manufacturer control box using the Amphenol connectors provided.
The oxygen sensor itself was then placed in a piece of exhaust pipe that was slid on the
end of the vehicle tailpipe. The oxygen sensor readings were checked in the monitor
mode to see if the equivalence ratio was near 1. For GM vehicles, the data were
3-34
-------
Figure 3-12
6-PARAMETER LOGGER IDENTIFICATION AND SENSOR VERIFICATION
!©!)3<333©Qr
Vehicle Sample No.: S B {get from Cover Sheet)
VIN: {Fill for 1981+cars)
Model Year: {get from label under the hood)
Make:
Model:
Transmission: Automatic Manual: 3 4 5 6
Odometer:
Vehicle Selection Parameters:
GM FORD CHRY MITS MAZD NISS TOYT
Year
Make
Model
Engine _
VIN ;
2/4 Wheel Drive ' ¦
S/D Cam _
Box Test -
Electronic Dash
Check Error Codes Before Installation:
Check Error Codes After Installation:
3-35
-------
downloaded to check the oxygen sensor readings. If the oxygen sensor readings were
found to be satisfactory, the vehicle was taken to the muffler shop for the boss
installation. After the installation, the oxygen sensor was installed into the boss and
connected to the datalogger system, and the vehicle exhaust system was tested to be free
of exhaust leaks. The vehicle was then taken on a short drive. All the vehicle
parameters were then examined to see if they were satisfactory.
After the Spokane installations and before the Baltimore installations
began, the oxygen sensor responses were checked against air and 0.1 percent oxygen in
nitrogen, which simulates an exhaust gas from an engine running near the stoichiometric
air/fuel ratio. The voltage from the oxygen control box was measured. The air calibra-
tion check was performed on 18 oxygen sensors that were available, and the 0.1 percent
oxygen check was performed on four oxygen sensors. One oxygen sensor had been sent
to Ford and another to Toyota, and they could not be checked before the vehicle
installations. Table 3-4 shows the logger number, oxygen sensor serial number, and the
voltages recorded for the oxygen sensors. According to data provided by NGK, the
oxygen sensor voltage for air should be between 4.25 and 4.75 volts, and the voltage for
oxygen concentration of 0.1 percent should be near 3.000 volts.
The next sections describe the installation process for each manufacturer.
3.2.2.1 GM Vehicle Instrumentation
Before proceeding with the GM installation, the GM datalogger's
interfacing with the vehicle was tested by directly connecting the GM datalogger to the
vehicle ALDL connector. If the communication was successful, the power cable was
installed to the fuse box ignition ON power. Next, the ALDL cable was connected to
the diagnostic connector under the steering wheel. The ground cable was installed next
to the power cable in front and also to the datalogger box in the trunk.
3-36
-------
Table 3-4
Oxygen Sensor Calibration Check
(Performed on March 4, 1992)
Datalogger
Box #
Manufacturer
Oxygen
Sensor #
Oxygen Sensor
Voltage with Air
Oxygen Sensor
Voltage with
0.1% Oxygen
Blue 5
Ford
1925-1228
4.354
3.003
Blue 15
Mitsubishi
1925-1222
4.278
3.007
Blue 6
Ford
1925-1224
4.312
3.006
Orange 21
Nissan
1925-1227
4.205
3.007
Orange 18
Nissan
1926-1226
4.226
-
Red 1
Chrysler
1925-1223
4.238
-
Red 2
Chrysler
1925-1221
4.263
-
Blue 4
Ford
1925-1220
4.286
-
Blue 3
Ford
1925-1216
4.226
-
Green 16
Mazda
1925-1217
4.321
-
Green 17
Mazda
1925-1219
4.286
-
Yellow 8
GM
1925-1231
4.234
-
Yellow 12
GM
1925-1213
4.226
-
Yellow 11
GM
1925-1212
4.228
-
Yellow 10
GM
1925-1211
4.426
-
Yellow 13
GM
1925-1214
4.243
-
Yellow 14
GM
1925-1215
4.278
-
Yellow 9
GM
1925-1232
4.233
-
Blue 7
Ford
1925
-
-
White 20
Toyota
1925
-
-
3-37
-------
Figure 3-13 shows the 6-Parameter Logger Installation Notes form. This
form was filled out after all the wiring was connected between the vehicle andthe
datalogger box. The locations of all the vehicle parts removed during the installation
were also noted.
The 6-parameter datalogger software check was performed by using the
6-Parameter Logger Calibration Sheet shown in Figure 3-14. The vehicle was driven at a
known speed over a fixed driving route. The time, speed, and RPM of the vehicle were
recorded and compared with the values stored in the datalogger. After the data had
been checked, the datalogger was initialized and the vehicle was given to the owner.
3.2.2.2 Mitsubishi Vehicle Instrumentation
The procedures described above in the GM section for vehicle
identification, sensor verification, wire installation and software check were followed.
The forms presented in Figure 3-12, 3-13, and 3-14 were used. The Mitsubishi
datalogger was a Campbell Scientific CRIO-based system. The primary connections were
to the ECU and the multi-use connector. The connection to the ECU provided 12 volts
of power to the Mitsubishi interface box, and the connection to the multi-use connector
provided speed, RPM, throttle position, and coolant temperature information to the
interface box. The MAP was measured by using a vacuum sensor mounted on the air
intake plenum in the engine compartment. Once the basic wiring was determined to be
working properly, the oxygen sensor was installed. After the oxygen sensor was
connected to the interface box, the final software check was performed by using the form
described in Figure 3-14. The vehicle was driven on a specified route and the 6-
parameters were monitored during the drive. After the values were confirmed to be
appropriate in both the monitoring mode as well as in the downloaded data, the vehicle
was given to the owner.
3-38
-------
Figure 3-13
6-PARAMETER LOGGER INSTALLATION NOTES
Datalogger System Color/Number
GM FORD CHRY MITS MAZD NISS TOYT
Connection Location:
Power
Ground
Speed
RPM
THR
Circle: MAP MAF LV8
CTS
Wire Routing:
Vehicle location of vehicle parts removed during the installation:
Oxygen Sensor Box should be ON
Oxygen sensor location: Upstream of Muffler Downstream of Muffler
3-39
-------
Figure 3-14
6-PARAMETER LOGGER CALIBRATION SHEET
Non-GM only: Check that memory modules are empty
Non-GM only: Check that FILL & STOP switch is ON.
Non-GM only: Name of program installed on CR10:
Non-GM only: Record values at idle:
Time
Speed
RPM
THR
Circle: MAP MAF LV8
CTS
PHI
Idle:
WOT:
GM and non-GM: Drive the vehicle at a known speed and record:
Measurement
Vehicle's
Instruments
Datalogger
Reading
Date
Time
Speed
RPM
Comments:
GM and Non-GM: Reset datalogger after test drive.
Non-GM only: Connect memory modules.
3-40
-------
3.2.2.3
Ford Vehicle Instrumentation
The procedures described in the above sections for vehicle instrumentation
were followed. All the Ford connections were made through a tee connector into the
ECU of the vehicle. The ECU was removed and the tee connector was put into its
place. The ECU was then connected to the tee harness. This was the only connection
and this harness was connected directly into the Ford manufacturer's box, which received
all the sensor information and power from this connection. After the vehicle was
instrumented and the idle mode test was performed, the oxygen sensor was installed.
The final software check was performed using the form described in Figure 3-14.
3.2.2.4 Chrysler Vehicle Instrumentation
All the Chrysler connections were made through a diagnostic connector,
which was found next to the battery in the engine compartment. The harnesses were
routed underneath the vehicle and up through the rubber grommet in the vehicle trunk.
This harness connected to the manufacturer interface box in the trunk. The box and the
oxygen sensor were then connected to each other and to the CR10. All the procedures
described in the above sections were followed before the vehicle was released to the
owner.
3.2.2.5 Mazda Vehicle Instrumentation
All the procedures discussed in the above sections were followed. The
Mazda-specific installation procedures required the connection of a manufacturer-
supplied ECU in place of the original vehicle ECU. The speed sensor cable, which was
connected to the speedometer cable, was used to determine the speed of the vehicle. All
the other sensors (MAP, RPM, coolant temperature, and throttle position) were
determined from the ECU. The oxygen sensor was directly connected to the
3-41
-------
manufacturer interface box. Depending on the particular model of the vehicle, the ECU
box had to be configured to a particular setting.
After the oxygen sensor was installed and the software check was
performed the vehicle was released to the owner.
3.2.2.6 Toyota Installation Procedures
The primary connection for the Toyota installation involved the
replacement of the original ECU with the manufacturer-supplied ECU box. All the
vehicle parameters were determined from this ECU box. Connections also had to be
made for the 12-volt power to the manufacturer interface box. Because of the large size
of the Toyota interface box, it was installed on the back seat of the vehicle. After all the
harnesses were connected, each signal had to be calibrated on the manufacturer-specific
box. The calibration process was required for both zero and span values. When this
calibration process was completed and the oxygen sensor was installed on the vehicle, it
was released to the owner.
3.2.2.7 Nissan Vehicle Instrumentation
The procedures specified for vehicle identification, sensor verification,
datalogger installation, and software check described in the sections above were
followed. Nissan provided a special ECU box with a model-specific EPROM chip that
had to be substituted for the original ECU. In addition, connectors were used to supply
12-volt power as well as a link to the diagnostic connector for Nissan vehicles. After the
initial wires were connected, the vehicle parameters were checked at idle. After the
proper values were monitored for all five vehicle parameters, the oxygen sensor was
installed, followed by the final 6-parameter software check. Before the vehicle was
released to the owner, the datalogger was reset so that only owner driving was collected
on the memory modules.
3-42
-------
3.3 Datalogger Removal Procedures
The dataloggers were removed from the vehicles during appointments with
vehicle owners. At the appointed time, a Radian van rendezvoused with the driver and
vehicle to remove the datalogger. To minimize the inconvenience to the owner of the
vehicle, owners were offered the option of having dataloggers removed at their house or
workplace, but most preferred to return to the I/M station to have the equipment
removed.
3.3.1 3-Parameter Logger Removal
First, the datalogger and speed transducer were checked for proper
operation before they were removed from the vehicle. This was done by connecting the
datalogger to the laptop PC as it was during the datalogger calibration. The datalogger
status flag was changed from RUN to CAL. The displayed date and time were checked.
The vehicle was then taken for a short drive to check the displayed vehicle speed on the
PC against the vehicle's speedometer. These two values were recorded on the
3-Parameter Logger Removal Report, shown in Figure 3-15. This check was made to
ensure that the speed transducer was still working or to determine if it had been
degraded in some way, for example, if some of the magnets had fallen off the rotating
shaft. Finally, during the drive a wide-open throttle acceleration was performed to aid
data analysis.
The datalogger was then carefully removed from the vehicle. While the
mechanic was removing the datalogger from the vehicle, the assistant discussed the
vehicle's usage with the owner to determine how the vehicle had been driven during the
week the datalogger was collecting data. The purpose of this the usage information form
(shown in Figure 3-16) was to determine if there was any reason to throw out the data
collected by the datalogger. Finally, the owner was paid the remainder of the incentive.
The data from the logger memory were downloaded to the PC hard disk. The file name
3-43
-------
Figure 3-15
3-PARAMETER LOGGER REMOVAL REPORT
Data Points Collected:
Put datalogger in CAL mode.
Odometer:
Datalogger
Reading
Date
Time (CST)
Speed (> 30 mph)
RPM at Fast Idle
Perform WOT acceleration from <10 mph to 35-60 mph. Local Time
Explanation of removal drive abnormalities:
Hit Alt X to exitPROCOM
Datalogger Number (get from logger case before removal!)
Remove datalogger from vehicle.
Measurement
Vehicle's
Instruments
3-44
-------
Figure 3-16
USAGE INFORMATION
IFteroatEwal) H©ac°]©r
During the last week:
Was the vehicle driven?
Did you go on any long trips?
Who drove the vehicle?
Did you have your car worked on?
3-45
-------
used was the last eight characters of the VIN. For reliability, downloading was done in
the hotel at the end of the day. To do the downloading, the datalogger removed from
the vehicle was powered by a 12-volt supply. Immediately after the data were
transferred onto the PC hard disk, the raw data file was compressed, and two copies on
floppies of the compressed hard disk file were made. One floppy was archived on site.
The second copy was mailed with all the other floppies from dataloggers retrieved that
day to Austin. The 3-Parameter Data Downloading Sheet, shown in Figure 3-17, was
filled out to document details of the data retrieval.
3.3.2 6-Parameter Logger Removal
For the GM dataloggers, the removal team confirmed that the datalogger
was in the data collecting mode before it was removed. Figure 3-18 shows the
6-Parameter Datalogger Removal Report. The cables and connectors were removed
from the vehicle and the vehicle was returned to its original state. The datalogger
installation notes were consulted to make sure that all vehicle parts were returned to
original condition.
For the non-GM dataloggers, the removal team first removed the memory
modules and connected the computer to the CR10 datalogger. They monitored the 6
parameters in idle mode and noted the readings on the form described in Figure 3-18.
In addition, the vehicle was taken on a short drive to note the speed and RPM of the
vehicle. The vehicle instrument readings were compared with the datalogger readings
during the drive. If all parameters read properly, all the cables were removed from the
vehicle and the battery was disconnected from the CR10.
For both the GM and the non-GM dataloggers, the oxygen sensor was
removed from the exhaust system and new exhaust system components after the catalytic
converter were installed.
3-46
-------
Figure 3-17
3-PARAMETER DATA DOWNLOADING SHEET
Datalogger Number (get from datalogger case!)
Dump data to PC hard disk.
Name the Raw Data file with the last 8 characters of the VIN:
• 3RD
Record the size of the *3RD file as seen on the hard disk:
Mbytes
Compress the file using PACK to create a new file.
Name the ComPressed file with the last 8 characters of the VIN:
• 3CP
Make a copy of this file on each of two 3.5" 1.44 Mbyte floppies.
Label each floppy with:
Filename*3CP
Vehicle Sample ID
Today's Date
Fed Ex one floppy to Austin.
Archive the other floppy on site.
3-47
-------
Figure 3-18
6-PARAMETER LOGGER REMOVAL REPORT
Odometer:
Datalogger System Color/Number:
GM:
(Get number from the system box, not the datasheet!)
Date:
Time (local):
Make sure PRODAS is in COLLECT mode.
Remove cables and GM datalogger.
Non-GM:
Counter:
Disconnect memory modules.
Record values at idle:
Speed
RPM
THR
Circle: MAP MAF LV8
CTS
PHI
Drive the vehicle and record:
Measurement
Vehicle's
Instruments
Date
Time (local)
Speed
RPM
Remove cables and non-GM datalogger from vehicle.
3-48
-------
The owner was also paid the balance of the incentive originally agreed on.
At this time, the Usage Information form shown in Figure 3-16 was also completed.
3.3.3 Data Handling
As data were downloaded in the hotel room from the dataloggers onto the
portable PC, the data were put in a SAS-compatible format for loading onto a
mainframe computer.
Each of the datalogger systems had software that read the logger's
memory, converted the information in memory into engineering units for each parameter
measured, and sent it to the portable PC. Thus, the format of the data in the
downloaded files on the PC could be controlled by modifying the downloading software.
Because the 3-parameter and 6-parameter loggers were measuring different parameters
in different ways, the downloaded data files were different. The downloaded data file
formats for the two types of dataloggers are described below.
3.3.3.1 3-Parameter Data Downloading
As the data were downloaded from the datalogger to the PC, the data were
automatically put into the SAS-compatible format, shown in Figure 3-19. The first
twenty-four lines in the downloaded data file give information about the vehicle to which
the data in the following seven columns apply. These columns of data in the
downloaded file provide the operational information about the vehicle.
These seven parameters are self-explanatory except for the last two. Status
denotes whether the data for the observation was taken while the datalogger was in the
calibration mode and the vehicle was being driven by the instrumentation staff or
whether the data were being obtained by the regular driver of the vehicle. Parity tells
whether the data on each observation line of the file passes a parity check. This is used
3-49
-------
Figure 3-19. 3-Parameter Downloaded Data Format
dd
Vehicle Identification :
Sample ID = bl3 0
Motor Vehicle ID = Ifabp44e2kf117098
Model Year = 19 89
Make = ford
Model = mustanglx
Transmission = automatic
Odometer Reading = 45277
Speed Calibration :
1192 pulses is equivalent to 800 feet
RPM Calibration :
Piston engine
Number of Cylinders = 8
Number of coils » 1
03/11/92 12:24:54
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Date Time Speed
(12/25/91) (HHMMSS) (m/s)
833
78.6
cal
1233
31.6
cal
667
37.3
cal
700
42.2
cal
700
47.8
cal
700
47.0
cal
700
47.0
cal
700
47.0
cal
700
47.0
cal
733
48.6
cal
733
48.6
cal
1000
55.9
cal
967
52.7
cal
867
49 .5
cal
733
47.8
cal
733
49.5
cal
800
48.6
cal
1433
71.3
cal
1433
65.7
cal
833
44.6
cal
667
48.6
cal
667
48.6
cal
RPM
MAP
Status
(rpm) (kPa) (RUN or (»or
CAL) parity_error)
3-50
-------
to greatly improve the reliability of the data on each line and can be used to spot errors
in the logged data. Other than this information, no other data were recorded in the
downloaded files.
3.3.3.2 6-Parameter Data Downloading
The final phase of the data handling process was downloading the
6-parameter data from the dataloggers. This was done by following the instructions on
the form described in Figures 3-20 and 3-21. For the non-GM dataloggers, the CR10
memory modules were connected to the PC through an appropriate interface. The
vehicle files were named so that the identity of the memory modules and the vehicle
sample number could be identified. The 6-parameter downloaded data have no values in
the file which specified the vehicle to which the data applied. All the files downloaded
for a particular vehicle for memory module 1 and memory module 2 were concatenated
to produce one file that contained all the data for that particular vehicle. The file was
converted to the final output format and was examined using an editor. Figure 3-22
shows the format of the downloaded 6-parameter data. Once a file was found to be
reasonable, it was compressed so that it could fit onto a 1.44 Mb floppy. Two copies of
the compressed data were made. One was sent to Austin and one was filed with the
data packet on site.
Figure 3-21 was used for downloading the files from the GM datalogger.
The output format of the file was available directly from the datalogger software. The
file was examined to make sure that all the data were available on the file. The data file
was compressed and two copies were made of the compressed data file.
3-51
-------
Figure 3-20
6-PARAMETER DATA DOWNLOADING SHEET
(GM only)
Datalogger System Color/Number
(Get number from the system box, not the datasheet!)
Read and Convert data using PRODAS.
Name the raw data file with the last 8 characters of the VIN:
• PRN
Examine the *PRN file.
Record »PRN file size in /PRODAS/DATA directory: bytes
Compress the *PRN file with GMPACK to make a *6CP file.
Make a copy of this file on each of two 3.5" 1.44 Mbyte floppies.
Label each floppy with:
Filename*6CP
Vehicle Sample ID
Today's Date
Initialize the GM datalogger.
Fed Ex one floppy to Austin.
Archive the other floppy on site.
3-52
-------
Figure 3-21
6-PARAMETER DATA DOWNLOADING SHEET
(non-GM)
Datalogger System Color/Number
(Get number from the system box, not the datasheet!)
Connect Memory Module 1 to PC with the SC 532 interface.
Collect all data files using SMCOM.
Name Memory Module 1 files with Vehicle Sample Number: 1
Repeat procedure for Memory Module 2. Name these 2 files.
Use CONCAT to concatenate the »DAT files into:
• DAT using the last eight VIN digits.
Use SPLIT to convert the *DAT file to a *PRN file. bytes
Examine the *PRN file with the SEE editor. Is it reasonable?
Use SIXPACK to compress the *PRN file into a «6CP file.
Make a copy of this file on each of two 3.5" 1.44 Mbyte floppies.
Label each floppy with:
Filename*6CP
Vehicle Sample ID
Today's Date
Reset memory modules.
Fed Ex one floppy to Austin.
Archive the other floppy on site.
3-53
-------
85
85
85
85
85
85
85
85
85
85
85
85
65
85
85
85
85
65
85
85
85
85
85
85
85
85
85
85
85
85
85
85
85
85
85
85
85
85
85
85
Figure 3-22
6-Parameter Downloaded Data Format
Time
Speed Throttle Coolant
HHMM SS (mph) RPM Position Load Temperature
1641
52
1
925
1
98
31
1. 023
1641
53
0
925
1
96
31
1.022
1641
54
0
925
1
98
31
1.022
1641
55
0
925
1
98
31
1.022
1641
56
0
925
1
97
31
1.024
1641
57
0
925
1
98
32
1.024
1641
58
0
925
1
96
32
1.023
1641
59
0
925
1
96
32
1.027
1642
0
0
925
1
96
32
1.001
1642
1
0
925
1
96
32
0.998
1642
2
0
900
1
94
33
0.995
1642
3
0
925
1
96
33
0.999
1642
4
0
900
1
96
34
1.002
1642
5
0
900
1
95
34
1.010
1642
6
0
925
1
94
34
1.011
1642
7
0
925
1
96
34
1.015
1642
8
0
875
1
93
34
1.018
1642
9
0
900
1
92
34
1.026
1642
10
0
900
1
95
34
1.010
1642
11
0
900
1
94
35
1.025
1642
12
0
900
1
94
35
1.016
1642
13
0
900
1
96
35
1.018
1642
14
0
925
1
95
35
1.027
1642
15
0
900
1
94
36
1.029
1642
16
0
900
1
94
36
1.017
1642
17
0
925
1
91
36
1.025
1642
18
3
1425
10
154
36
0.999
1642
19
8
1625
15
161
36
0.993
1642
20
12
1825
21
180
37
1.001
1642
21
17
2250
26
189
37
0.975
1642
22
22
2000
26
205
37
1.001
1642
23
26
2075
26
196
37
1.002
1642
24
29
2100
21
168
37
1.048
1642
25
30
1625
13
142
37
1.067
1642
26
30
1225
1
56
37
1.051
1642
27
29
1125
1
50
38
1.196
1S42
28
27
900
1
46
38
1.251
1642
29
25
950
1
55
38
1.277
1642
30
22
775
1
65
38
1.287
1642
31
17
750
1
82
39
1.155
3-54
-------
4.0 FIELD RESULTS
The results of the field efforts in Spokane and Baltimore are presented in
this section. First, a review of the reasons that EPA chose Spokane and Baltimore for
the driving pattern study are reviewed. Next, a summary of the facilities at the
inspection/maintenance stations and the activities of the installation and removal staff
are presented. The final part of this section reviews the success rates for solicitation and
datalogger instrumentation for the 3- and 6-parameter dataloggers. Tables are presented
which show the identity of the vehicles that were solicited, as well as those that were
instrumented in this study.
4.1 Selection of Cities
EPA selected two cities, Spokane, Washington and Baltimore, Maryland,
for the driving modes study. The selection was based on a number of factors. For the
instrumented vehicle study, the primary factor was the presence of a centralized
inspection/maintenance (I/M) program in the two cities. A centralized I/M station
provided a convenient location to sample vehicles representatively in a situation where
drivers would be aware of exhaust emissions of their vehicles. Driving in Spokane
consists of city driving on flat as well as hilly terrain, and Baltimore presented a variety
of city driving conditions, both on city streets and on the highway. The I/M program in
the State of Maryland uses 10 locations all over the state for testing vehicles. Two
locations, one in the Baltimore city area and one in the Baltimore county area,
represented by Locations 3 and 4, were used in this study. The Spokane I/M program is
conducted through only one location, which was used in this study.
The desirable characteristics of the survey sites were considered before the
cities were chosen for the measurement of driving patterns. The characteristics
considered were:
4-1
-------
• Nonattainment area,
• Centralized I/M;
• Transportation Planning System Model;
• Terrain;
• Altitude;
• Precipitation;
• Road System; and
Traffic.
It was appropriate to use an emissions nonattainment area for the study;
however, the question arose whether ozone nonattainment areas, CO nonattainment
areas, or areas which attain neither the CO nor the ozone NAAQS should have been
considered. No reason could be given as to how driving in an attainment or a
nonattainment area would affect driving behavior.
From a practical point of view, the survey site that had centralized I/M
made the procurement of instrumented vehicles easier. For the chase car study, a survey
site with a good urban transportation planning system model was needed.
The terrain and altitude of a survey site affects the way engines operate
and vehicles are driven. Since most vehicles are driven at low altitudes, it seemed
appropriate to choose low-altitude survey sites. Since the presence of many steep hills
can affect vehicle driving patterns, and most large cities are relatively flat, it was
appropriate to chose a city which is relatively flat.
4-2
-------
The presence of precipitation on the road may affect driving patterns, so
cities expected to have a large amount of precipitation during the period of the study
were ranked lower.
The characteristics of the traffic and road systems of the survey site may
greatly affect driving patterns. The driving characteristics in a city that has congested
traffic will be significantly different from those in a city that has relatively open traffic.
Trips in large, spread-out cities will tend to be longer than those in small, compact cities.
For example, driving patterns in downtown Boston would be different than in Los
Angeles.
Table 4-1 lists the nonattainment areas within the continental United
States that have centralized I/M programs. The table contains information about the
I/M program, the altitude, and precipitation in these 25 areas. This information, along
with other information specific to the collection of data by the chase car technique, was
considered in the selection of candidate cities for the measurement of vehicle driving
patterns.
All 25 areas use centralized I/M programs; however, only 20 use
centralized I/M exclusively. For example, in New Jersey vehicles may be inspected at
the centralized state station, or for a higher fee they may be inspected at decentralized
stations. While most vehicles go to the state stations, it is possible that the vehicles
inspected at the decentralized stations are driven differently than those inspected at the
centralized stations. Consequently, the five areas that have a combination of centralized
and decentralized I/M (New Jersey; Washington, DC; Jacksonville, Florida; Miami,
Florida; Tampa, Florida) were less attractive as candidate sites than those that have only
centralized inspection.
All of the sites except Tucson, Arizona and Spokane, Washington have
altitudes below 1300 feet. The higher altitudes of these two cities may cause different
-------
Table 4-1
Attributes of Nonattainment Areas with Centralized I/M Programs
ttonaitalnnwil
Modal Years of
t/M Program Type
Altitude
Average Per Cent
Average
Average Number
Comments
Ai«b
Light-Duty Vehicles
of Days
Monthly
Tested for Emissions
With > 0.01 fneh
:: Precipitation
With > 11nch Snow
Precipitation
(C = Centralized}
{feel)
(%)
(inches)
(0 i Decentralized}
(OCT NOV DEC JAN FEB)
STATES
Connecticut (e.g. Hartford)
68+
C
169
36
3.7
01433
Frequent snow
Delaware (e.g. Wilmington)
68+
C
74
32
3.2
01122
*
New Jersey
All
C + D
160
34
3.4
00222
Partly decentralized l/M
METROPOLITAN AREAS
Baltimore MD
79+
C
148
32
3.1
00 122
•
Washington suburbs MD
79+
C
10
32
2.9
00121
Suburbs
Washington DC
AH
C + D
10
32
2.9
00121
Partly decentralized l/M
Jacksonville FL
75+
C + D
26
25
29
00000
Partly decentralized l/M
Miami FL
75+
C + D
7
28
3.2
00000
Partly decentralized l/M
Tampa FL
75+
C + D
19
20
23
00000
Partly decentralized l/M
Memphis TN
All
C
258
30
4.1
00111
•
Nashville TN
79+
C
590
34
3.8
00111
•
Cleveland OH
75+
C
777
46
26
02444
Frequent precip, (requent snow
Louisville KY
All
C
477
35
3.2
00121
•
Louisville suburbs IN
76+
C
477
35
3.2
00121
Suburbs
Chicago suburbs IN
76+
C
801
34
2.4
01333
Frequent snow
Chicago IL
68+
C
658
32
1.9
01333
Frequent snow
E.Sl.Louis IL
68+
C
535
30
2.2
01122
Suburbs
Milwaukee Wl
76+
C
672
33
1.8
01343
Frequent snow
Minneapolis MN
76+
C
834
22
1.2
02333
Frequent snow
Phoenix AZ
67+
C
1112
12
0.6
00000
•
Tucson AZ
67+
C
2584
13
0.8
01 000
Higher altitude
Medford OR
71 +
C
1298
53
2.7
0 1111
Frequent precip
Portland OR
71 +
C
21
59
5.0
00111
Frequent precip
Seattle WA
68+
C
19
63
5.1
01121
Frequent precip
Spokane WA
68+
C
2357
43
2.0
02563
Higher altitude, frequent precip,
freauent snow
-------
driving behavior because vehicles will not respond to driver demands as do the majority
of vehicles operated at lower altitudes. Also, the terrain of the 25 areas was considered
in the final selection of survey sites.
Precipitation can produce differences in measured driving patterns by
allowing wheel spin and by inducing drivers to drive more carefully than they would on
dry pavement. On the other hand, since wet or snowy roads are part of the environment
in which most vehicles are occasionally driven, it was reasonable to measure driving
behavior under wet conditions during part of the study; however, because the data in this
study were collected in winter, and the results must be relevant to year-round driving, an
excessive number of wet days was to be avoided.
Data collection was expected to occur sometime between October 1991
and February 1992. The average precipitation for the months of October through
February at each of the 25 areas is shown in Table 4-1 in terms of the average
percentage of days that have measurable precipitation, the average monthly precipitation
in equivalent inches of water, and the monthly average number of days where snowfall is
more than one inch.
Consideration of the percentage of days with precipitation showed that the
list can be broken into three groups: 12-22%, 30-36%, and 43-63%. The amount of
precipitation is also shown in the table. Large amounts of rainfall (in comparison with
small amounts) were not expected to be a problem because roads will drain and dry;
however, days with snow could be a problem, especially in cold climates, because roads
could remain slippery for days after the snowfall. Accordingly, the table also shows the
average number of days with snowfall by month. In general, the number of days with
snowfall is low in October and November, but is high in December, January, and
February.
4-5
-------
An additional, although perhaps minor, consideration made for either type
of study was the presence of a nearby large city that could attract vehicles from the area
of interest and thus increase the amount of interstate driving over what would otherwise
occur. Baltimore, Maiyland was expected to be large enough to have driving patterns
like those of cities rather than those of suburbs, although other large cities are nearby.
After taking all considerations into account, EPA selected Spokane and
Baltimore. Auxiliary driving pattern studies are being carried out in Los Angeles and
Atlanta. The California Air Resources Board is sponsoring a chase car study in Los
Angeles. EPA's Office of Research and Development is sponsoring a 3-parameter
instrumented vehicle study in Atlanta. The addition of these cities should provide a
proper regional representation.
4.2 Activities at Inspection/Maintenance Facilities
In both cities, the project staff consisted of a solicitor, two 3-parameter
installers, two 6-parameter installers, and two datalogger removers. However, the staff
shared responsibilities as the need arose.
Physical installation and calibration of the 3-parameter datalogger took
from 40 minutes to 2 hours, depending on the complexity of the application. The
average installation time was about 70 minutes. In most cases, the vehicle owner waited
in the station for the vehicle to be instrumented. Installation of the 6-parameter systems
took from 1-1/2 to 4 hours. This time included a trip to a nearby muffler shop for
welding the oxygen sensor boss on the exhaust system. The average 6-parameter system
installation time was about 2 and 1/2 hours.
Instrument removals began one week after instrumentations began. All but
eight of the 3-parameter removals and all of the 6-parameter removals took place at the
I/M stations. 3-parameter removals took about 30 minutes per vehicle and 6-parameter
-------
removals took approximately two hours per vehicle, including the trip to and from the
muffler shop to replace the back portion of the vehicle's exhaust system.
Instrumentations in Spokane began on Monday, February 3 and continued
through February 22. Since the station was closed for holiday on February 17,
instrumentations were performed for a total of 17 days. The removal team, which began
its operations on February 10, continued to work through March 2.
Instrumentations in Baltimore began at the Baltimore County station on
Thursday, March 5 and continued through Monday, March 16. During that afternoon
equipment and personnel were moved to the Baltimore City station. Installations began
the next day and ended on Thursday, March 26. Instrumentations were performed for a
total of 19 days. The removal team, which began its operation on March 12, continued
to work through April 5.
4.2.1 Spokane, Washington
The instrumentation and removal of 3- and 6-parameter dataloggers took
place at the Inspection and Maintenance (I/M) station at South 211 Howe Street. The
station is located about four miles east of downtown Spokane in a commercial and
residential area. Eligible vehicles with registration addresses having 992 as the first three
digits of the zip code must be inspected. This station is the only I/M facility serving
greater Spokane and is run by VI'll, Inc.; therefore, most of the eligible vehicles
populating Spokane's roadways must obtain inspections at this facility once every two
years. Vehicles of commuters living 20 miles east in Coeur d'Alene, Idaho are not
inspected at this facility.
A biennial system is used to determine which vehicles are inspected. Even
model-year vehicles are inspected in even years. Since this study was performed in 1992,
almost all vehicles that came through the I/M facility were of even model years. Since
-------
new vehicles are exempt from inspection for the first year, the newest vehicle coming
through the inspection lane was the 1990 model year.
Figures 4-1 and 4-2 show the layout of the Spokane site. The I/M station
has four lanes, three of which were used for state inspections during the instrumentation
phase of the project. The fourth lane was used solely to install the instrumentation. The
station is operated from 9:00 AM to 6:00 PM on Monday, Tuesday, Thursday, and Friday
and until 8:00 PM on Wednesday. The station was open from 9:00 AM to 1:00 PM on
Saturdays and was closed on Sundays. The instrumentation and removal staff solicited
vehicles during all station hours.
The I/M program in the Spokane area is a centralized registration-
enforced program. It is a biennial program conducted by using the idle test for pre-1981
vehicles and the two-speed (idle plus 2500 rpm) test for the 1981+ vehicles. There is no
tampering inspection. The fee for the test is $16 and the cost waiver requirements are
$50 for the pre-1981 vehicles and $150 for 1981+ vehicles. All 1968 and newer vehicles
must be tested. Diesel vehicles, motorcycles, and alternatively fueled vehicles are
exempt. Vehicles are run at the desired rpm levels, during which the exhaust emissions
are tested. The cutpoints for all vehicle types for the I/M test are:
Model Year
HC (ppm)
CO (%)
1968-1974
1000
6.0
1975-1980
600
3.0
1981+ no/cat
600
3.0
1981+ w/cat
220
1.2
4-8
-------
NORTH
PSD'
230
\r
335'
HOVE." STREET
SPOKANE SKI
Figure 4-1. Howe Street Spokane Site
4-9
-------
)
-------
4.2.2
Baltimore, Maryland
Maryland inspection stations 3 and 4 were used as locations for recruiting
vehicles for the driving modes study. Station 3 is at 1001 Exeter Hall Avenue,
Baltimore, Maryland 21218. Station 4 is at 7969 Rossville Boulevard, Baltimore,
Maryland 21236. Location 3 is in Baltimore city and Location 4 is in Baltimore county.
Station 4 has the highest volume in the Maryland I/M program. Figures 4-3 and 4-4
show the layout of the lanes at the two sites. As can be seen, at both sites the vehicles
enter from a single lane or single entrance and move to the different lanes for the I/M
inspection.
The Maryland program is a biennial centralized I/M program. All vehicles
up to 26,000 lbs. that use gasoline fuel are tested. The hours of operation of the two
stations are from 8:00 a.m. to 6:00 p.m. Monday through Friday and 8:00 a.m. to 12:00
noon on Saturday. All 1977 and newer vehicles are covered in this program. Each
vehicle is assigned an inspection month as part of the two-year cycle. An inspection
notice is sent roughly 30 days before the inspection date for the vehicle. The owner has
until the end of the month to bring the vehicle in for inspection. For example, a person
who receives a notice on September 1 must have his vehicle inspected by the end of
October. The driver can bring in the vehicle at any time during the month of September
or October.
The I/M test consists of a visual test for the presence of a catalytic
converter, inspection and measurement of the fuel filler inlet restrictor, and
the exhaust gas emissions measurement. The exhaust gas measurement is done at idle
rpm only. The cutpoints for the I/M test are:
4-11
-------
Figure 4-3. Baltimore City I/M Station
4-12
-------
Figure 4-4. Baltimore County I/M Station
4-13
-------
Vehicles Under 10,000 lbs
Model Year
HC (ppm)
CO {%)
1977
500
6.0
1978
430
5.5
1979
400
4.0
1980
220
1.7
1981 +
220
1.2
4.3 Vehicle Sampling
The private vehicles to be instrumented in the study were selected from the
vehicles that passed through the I/M station at the time of the study. This section
describes the vehicles that were instrumented and also those vehicles solicited for
participation in the study but that were not instrumented. The method of vehicle
selection is also reviewed.
4.3.1 Summary of Vehicles Solicited
Table 4-2 gives a summary of the 247 vehicles solicited in Spokane
and the 483 vehicles solicited in Baltimore. The table presents the vehicles in the order
in which they were approached at the I/M station.
The sample number in Table 4-2 was assigned to every vehicle
approached by the solicitor. The sample numbers are unique to each vehicle and were
assigned to a vehicle whether or not the vehicle was instrumented. The sample number
begins with an "S" to indicate the sample applies to Spokane or with a "B" to indicate the
sample applies to Baltimore. Vehicles with sample numbers B001 to B257 were solicited
at the Baltimore County station. All other Baltimore vehicles were solicited at the
4-14
-------
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Number
Candidate
Method
Age
Type
Origin
Number
S 001
3
T
M
2
D
ELA-068
S 002
3
S001
M
2
D
73417-7
S 003
6
Q
M
1
F
215-DRE
S 004
3
T
M
1
F
342-DYA
S 005
3
T
M
1
F
151-CYB
S 006
6
Q
M
1
0
463-DKS
S 007
6
Q
O
1
0
628-DKU
S 008
3
T
Y
1
D
769-BOJ
S 009
3
T
M
1
D
966-DKS
S 010
3
S009
M
1
D
LVH-461
S 011
3
S009
M
1
D
CN 4991 (IL)
S 012
6
Q
O
1
D
757-CYD
S 013
6
0
O
1
D
158-DRE
S 014
6
Q
M
1
D
O40-CYD
S 015
3
T
M
1
0
SLH-930
S 016
6
Q
M
1
F
238-CMT
S 017
3
T
M
1
0
WCH 049
S 018
3
S017
M
1
D
HQD-876
S 019
6
Q
M
1
F
270-CYG
S 020
3
T
M
1
F
5463(MT)
S 021
3
S020
M
1
F
HMH-681
S 022
3
S020
M
1
F
157-BNA
S 023
6
Q
M
1
F
737-CYD
S 024
3
S020
M
1
F
597-DEV
S 025
3
T
Y
F
73829-T
S 026
3
T
M
1
0
C2D-228
S 027
3
S026
M
1
D
017-AFN
S 028
6
Q
O
1
D
409-CYB
S 029
3
T
Y
1
F
HZT-048
S 030
3
S029
Y
1
F
CSJ-604
S 031
3
T
M
1
D
NCY-333
S 032
3
S031
M
1
D
897-DRG
S 033
6
Q
M
1
D
S 034
3
T
M
1
D
321-BFT
S 035
3
T
O
1
D
009-DRM
S 036
3
T
M
F
DRPORK
S 037
6
Q
O
1
D
463-DRF
S 038
3
S034
M
1
D
WCH-461
S 039
3
S035
O
1
D
633-BFR
S 040
6
0
O
1
D
038-DRG
S 041
3
T
M
1
F
195-CDW
S 042
6
Q
M
1
D
729-CMT
S 043
3
T
M
2
D
73878T
Table 4-2
Solicitation Summary
hide
Driver
Logger
Distance
Logger
VIN
VIN
Final
;reen
Agrees
Installed
Sensor
Number
Conlirmed
Odomete
by Decoder
NA
NA
NA
NA
E14BHAH3213
NA
Y
Y
S
120
1GCGK24M8GJ162254
54239
Y
Y
NA
ORG18
JN1HJ01 P8LT366371
41290
Y
Y
S
121
KMHLA31J8HU121264
64011
Y
Y
M
103
WAUFC58AXLA061766
21809
Y
Y
NA
YEL10
1G2AF54T9L6235376
46827
Y
Y
NA
Yai4
2G1WL54RXL1169676
22991
Y
Y
W
123
1FAPP36X7JK162901
65465
NA
NA
NA
NA
1G2WK14W0JF285519
NA
NA
NA
NA
NA
1G1AW19W4G6199697
NA
Y
Y
W
142
1G3NT69U9GM372803
45232
N
NA
NA
NA
1P3XA5635LF774787
NA
N
NA
NA
NA
1G4NV54U0LMO79776
NA
Y
Y
NA
YEL9
3G4AL54N3LS605881
18526
Y
Y
S
139
HL41G8B367586
83282
Y
N1
NA
NA
JM1GD2226L1812401
NA
N
NA
NA
NA
1G3NT27U4GM376852
NA
Y
Y
M
111
GCFBAD246990
No
7877
Y
Y
NA
GRN16
JM1BG2265L0111262
N
NA
NA
NA
NA
N
NA
NA
NA
1G1AW27XXE6802201
NA
NA
NA
NA
NA
1HGCA6280JA000240
NA
Y
N2
NA
GRN17
JM1GD2227L1806476
NA
Y
Y
W
124
JF2KA83A2JD727208
50355
Y
Y
W
105
JT4RN50R6J0353266
55206
NA
NA
NA
NA
A6A057A339724
NA
Y
Y
S
133
39128622
No
71084
Y
Y
NA
YEL8
1G2HX54C9L1240876
22280
Y
N3
NA
NA
WBA-4007548
NA
Y
Y
S
104
LB110-B00914
73371
NA
NA
NA
NA
NA
N
NA
NA
NA
KL2TN5462LB317977
NA
N
NA
NA
NA
1C3XJ41K1LG431532
NA
Y
N2
NA
NA
2G2AF81R4J9232840
NA
N
NA
NA
NA
1G4HR54C9JH426908
NA
Y
Y
108
JT3FJ62G5J0087906
26642
NA
NA
NA
NA
1C3XC66R9LD898601
NA
Y
Y
W
136
1G4AT27P2EK537437
74264
Y
N2
NA
NA
2G3AJ51R5J9306520
NA
P
Y
Y
NA
YEL12
1G3AM54N2L6363462
21252
P
Y
Y
S
141
JT2AE92W1J3141485
45352
P
Y
Y
NA
BLU4
2FAPP37X8LB105148
17971
P
Y
Y
W
113
1B7FN14C7JS796743
78526
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logge
Number
Candidate
Method
Age
Typo
Origin
Number
Screen
Agrees
Install
S 044
6
Q
M
1
D
968-CMK
P
N
NA
S 045
6
O
O
1
F
856-CMK
P
Y
Y
S 046
3
T
M
O
UJ-5074
P
Y
Y
S 047
6
Q
M
1
0
455-DRG
P
Y
Y
S 048
3
T
O
D
117-BFT
P
Y
Y
S 049
6
O
O
1
D
P
N
NA
S 050
3
S031
M
1
D
472-AGB
P
Y
Y
S 051
6
Q
O
1
F
926-CMT
P
Y
Y
S 052
6
Q
M
1
D
215-CM0
P
N
NA
S 053
6
Q
M
1
D
190-CYA
P
Y
Y
S 054
3
T
M
D
14654-U
P
Y
Y
S 055
3
S035
O
1
D
PCD-307
P
Y
Y
S 056
6
Q
M
1
F
292-DRG
P
Y
N2
S 067
6
a
O
1
D
280-DRE
P
Y
N5
S 058
6
Q
O
1
D
345-D41
P
Y
Y
S 059
3
T
M
1
D
HZZ-695
P
N
NA
S 060
3
S059
M
1
P
N
NA
S 061
3
T
M
1
F
357-AIX
P
Y
Y
S 062
6
O
M
1
D
31100(ID)
P
Y
N1
S 063
6
Q
O
1
F
945-CYB
P
Y
Y
S 064
3
SO 59
M
1
D
865-CED
F
NA
NA
S 065
6
Q
O
1
F
202-BNB
P
NA
N8
S 066
3
T
M
1
D
996-DEV
P
Y
Y
S 067
3
SO 59
M
1
D
297-AFY
P
Y
Y
S 068
6
Q
M
F
666-CYD
P
Y
N2
S 069
3
T
y
1
F
501-DRF
P
Y
Y
S 070
6
Q
o
1
F
7877
P
Y
Y
S 071
6
a
o
1
D
650-DRF
P
N
NA
S 072
3
T
M
1
D
177-CEB
P
Y
Y
S 073
3
T
M
1
F
966-CLQ
P
Y
Y
S 074
6
Q
o
1
D
599-DKL
P
Y
N2
S 075
3
T
M
1
D
HMM-404
P
N
NA
S 076
3
S075
M
1
D
QLM-026
P
Y
Y
S 077
6
o
M
1
D
533-CMK
F
NA
NA
S 078
3
SO 59
M
1
0
RVL-049
P
Y
Y
S 079
3
T
M
1
D
734-CMK
P
Y
Y
S 080
6
Q
M
D
756-DKS
P
N
NA
S 081
6
Q
M
1
D
225PK361(CA)
F
NA
NA
S 082
6
Q
M
1
D
763-DRG
P
Y
Y
S 083
3
T
M
1
F
188-CMT
P
N
NA
$ 084
6
O
M
1
D
502-CMP
P
N
NA
S 085
3
S083
M
1
F
KZZ-703
F
NA
NA
S 086
3
T
M
2
D
53523-R
P
Y
N3
Distance
Sensor
Logger VIN
Number
VIN Final
Confirmed Odometer
by Decoder
NA
NA
1FAPP93J0LW139667
NA
NA
GRN16
1YV GD22B9L5241604
5037
M
106
2FTEF14N6GCA49786
12873
NA
RED1
1B3XA4636LF914826
26298
5
119
AA0BE8S328524
20978
NA
NA
NA
W
130
1MEBP89C9EG646134
26759
NA
GRN17
JM1GD222XL1822669
25800
NA
NA
1G1LT54G6LE180304
NA
NA
BLU6
1FACP52U5LG167445
30663
S
107
1B7HW14T0GS016645
83040
S
116
1J089AA196188
47804
NA
NA
J A3C R46 V2LZ041788
NA
NA
NA
1B3XC5639LD718780
NA
NA
BLU5
1FAC P50UXLG149557
9960
NA
NA
8W82L268781
NA
NA
NA
NA
S
127
JM1BF2227G0145368
68945
NA
NA
1G1LT54G8LY249993
NA
NA
WHT19
JT2SV21E2L0336391
29938
NA
NA
2FABP43F7GX134162
NA
NA
NA
JT2AE92E6J3034861
NA
M
145
1FABP3934EG169705
15865
S
146
1C3BH58E1GN168729
40911
NA
NA
JT3VN39WXL0023696
NA
S
135
JM1BF2327J0116975
65091
NA
BLU15
JA3CR46V8LZ036160
29246
NA
NA
1C3XC66R5LD858645
NA
128
2G2AK37H5E2268581
64645
M
126
JF2AN55B4GD439011
108311
NA
NA
1LNCM92EXLY703010
NA
NA
NA
1G1AZ37G9ER174613
NA
W
114
1FABP22X2GK155745
96676
NA
NA
2FACP73F4LX149931
NA
M
131
1C3BA54E0EG237812
S
122
1B3BZ18C5GD231063
62889
NA
NA
1P4FH44ROLX303404
NA
NA
NA
1FTCR14X1LPA46236
No
NA
NA
BLU7
1 MECM50U6LG&49782
30183
NA
NA
WAUFB0856EA072906
NA
NA
NA
1B3XP48K1LN166561
NA
NA
NA
JF2AM53B7CE437554
NA
m
NA
1FTCR15X1LPA46236
NA
-------
Final
OdomeK
NA
32133
25794
28543
NA
31368
94528
43014
NA
35663
32755
17629
NA
8983
NA
NA
42696
NA
05422
NA
23618
80682
25039
51643
NA
33451
NA
61203
NA
NA
15625
77522
21759
31085
NA
23397
NA
NA
33371
24701
23816
56293
27697
Table 4-2 (Continued)
.ogger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
indidale
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
3
S086
M
2
D
018-DRE
P
N
NA
NA
NA
1GNCT18Z2L8118030
3
S083
M
1
F
202-BNB
P
Y
Y
S
SN03
JT2AE92E6J3034861
6
a
M
2
O
807-DKU
P
Y
Y
NA
RE OS
1B4FK44R0LX29552O
3
S0B6
M
2
D
281-CMO
P
Y
Y
W
129
1GNDL15Z6LB143279
3
T
P
N
NA
NA
NA
2G1WN54TXL1156917
6
Q
1
D
591-CMT
P
Y
Y
NA
REDE
1P3XA46K8LF815124
6
Q
M
1
D
983-ALY
P
Y
Y
NA
YEL11
2G1WN54T4L9218037
3
T
M
1
D
EBN-30J(NJ)
Y
Y
W
109
1G2WK14W3JF252210
6
Q
M
1
D
703-DKU
P
Y
N6
NA
NA
1G3WS54T9LD395198
3
T
M
1
D
019-BFQ
P
Y
Y
M
110
1MEBP75X7GK641646
6
Q
M
D
596-DKP
P
Y
Y
NA
RED1
2B4FK4530LR632793
6
Q
M
1
D
289-CYB
P
Y
Y
NA
Yai3
1G4CW54C8L1624667
3
T
M
1
D
ENR-837
P
N
NA
NA
NA
1G1AX08X3CT128925
3
S099
M
1
D
905-CYE
P
Y
Y
132
1G1AW19R0G6112873
6
O
M
1
D
992-CMK
P
Y
N3
NA
NA
1FAPP959XLW190029
3
T
M
1
D
192-CYA
N
NA
NA
NA
1FAPP9597LW149311
3
SI 02
M
1
D
760-BVL
P
Y
Y
112
1G1LV14W9JE673479
3
T
M
1
D
581-CMG
N
NA
NA
NA
JE3CU36X4KU069206
3
S104
M
1
D
ELB-184
P
Y
Y
M
SNOB
1FABP0521CW107600
3
T
M
F
TU-2543
P
N
NA
NA
NA
RA32034992
6
O
1
D
012-CYH
P
Y
Y
NA
RED1
1B3XA4639LF838812
3
S106
M
F
UN-1035
P
Y
Y
S
143
JN6ND16Y0GW003899
6
Q
M
1
F
629-CMO
P
Y
Y
NA
WHT19
JT2SV24E2L3411553
3
T
M
D
653-A1Y
P
Y
Y
S
SN07
JB3BA26KXGU135565
6
Q
O
1
D
760-CYB
P
Y
N6
NA
NA
1G3WS54TXLD375851
3
T
M
1
F
235-CYE
P
Y
Y
W
144
1Y VG D31BXL5201264
6
Q
O
1
D
436 CYG
P
N
NA
NA
NA
3
T
M
1
F
440-ADF
P
Y
Y
S
118
KMHLF21J9HU104162
3
T
M
1
F
(TEMPORARY)
P
Y
N3
w
NA
JA3C U26X3LU017646
3
S115
M
1
F
259-CMS
P
Y
N3
NA
JHMCB7652LC072385
3
S116
M
1
F
962-DKU
P
Y
Y
w
125
JT2AE94A4L3346545
3
T
M
D
N7KQR
P
Y
Y
M
SN06
1G8CT18B1E0141922
6
Q
M
1
F
953-CMK
P
Y
Y
NA
ORG18
1N4GB22B6LC764253
3
T
M
1
F
090-DKR
P
Y
Y
W
139
JT2SV22EOL3407469
3
T
M
1
F
713-CB0
F
NA
NA
NA
NA
AL10073510
3
S121
M
1
F
345-CMK
P
Y
Y
M
106
JHMSM5423BC116394
6
O
M
1
D
265-CMS
P
NA
N6
NA
NA
1G3WH54T6LD351674
6
Q
M
1
O
014-CU1
F
NA
NA
NA
NA
JHMCB7662LC076249
6
Q
M
1
D
473-DRF
P
Y
Y
NA
BLU3
1LNLM9848LY702663
3
T
M
1
D
631-DKP
P
Y
Y
W
102
1Y1SK5467LZ153033
6
Q
O
1
D
832-CYD
P
Y
Y
NA
YEL14
1G3AJ54N1L6343342
3
T
M
1
F
166-DRH
P
Y
Y
S
SN04
1NXAE82G3JZ543668
6
Q
M
1
F
939-CMT
P
Y
Y
NA
GRN16
1Y VGD22B6L5224064
-------
Final
idometer
51749
14829
47809
17511
44227
34422
NA
62499
80173
NA
NA
NA
NA
25652
35167
NA
81461
104786
22519
NA
26395
NA
53668
NA
76753
NA
46113
NA
68905
NA
52065
NA
NA
98B50
25283
21799
NA
29650
NA
62257
NA
NA
59394
Table 4-2 (Continued)
ogger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
ndidale
Melhod
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
3
T
M
1
D
HMH-182
P
Y
Y
M
117
1FABP0757EW169481
6
Q
M
1
D
677-C4D
P
Y
Y
NA
BLU6
1MECM5344LG630501
3
T
Y
1
F
828-CMR
P
Y
Y
M
139
JHMAF5332ES007287
6
O
O
1
D
244-DRE
P
Y
Y
NA
YEL9
1G2NE54D1LC373895
3
T
M
1
F
671-BEX
P
Y
Y
W
121
1HGCA554XJA105600
3
T
M
2
D
34713-J
P
Y
Y
S
123
JB7FM55E2JP049855
3
T
M
3
D
881-BNB
P
N
NA
NA
NA
1FABP42E9JF199672
3
S136
M
3
D
721-CLJ
P
Y
Y
W
103
1G2PG9793GP263091
3
T
O
1
D
HMN-685
P
Y
Y
W
105
3M69FAR476562
3
T
M
2
F
44368-H
P
N
MA
NA
NA
JM2UC2213E0816360
3
S139
M
2
F
74007-T
P
N
NA
NA
NA
JN6ND06S6GW105228
3
S139
M
2
F
035-7UZ
P
N
NA
NA
NA
JA7FL24W0LPO20527
3
T
M
1
F
845-DKP
P
N
NA
NA
NA
2HGED634XLH529405
3
S142
M
1
F
840-BND
P
Y
Y
S
142
JM1BF2321J0174161
3
T
M
D
GQA-436
P
Y
Y
M
SN05
1FABP28A4DF128103
6
Q
M
1
D
460-CYA
P
N
NA
NA
NA
1FAC052U1LG214289
3
T
M
1
D
HMH-154
P
Y
Y
M
120
1FABP4634EH101344
3
S139
M
F
3191 0-J
P
Y
Y
S
113
JM2UF1133J0354042
6
Q
O
1
D
898-ORG
P
Y
Y
NA
YEL12
2G1WL54T8L1197259
6
Q
O
1
D
978-CMR
P
N
NA
NA
NA
3G4AH54NXLS613221
6
Q
O
1
D
619-DRF
P
Y
Y
NA
BLU4
1FAC P52UXLG233424
3
T
M
1
D
HMN-472
P
N
NA
NA
NA
1FABP4637EH151073
3
S151
M
1
D
WCH-040
P
Y
Y
W
111
1G3CX69B6G1323859
6
Q
M
1
F
145-DYR
P
Y
N7
NA
NA
JT2VV22W8L0101238
3
T
M
1
F
63B-AFV
P
Y
Y
M
124
SMG-2023538
3
T
Y
1
D
362-DC P
F
NA
NA
NA
NA
1ZVPT20C5L5206320
3
S155
Y
1
D
223-DKP
P
Y
Y
S
136
1ZVPT20C1L5138646
6
O
M
1
D
4P-4322A
P
N
NA
NA
NA
2G1WN54T2L1150707
3
T
Y
1
D
428-BVH
P
Y
Y
M
124
1B3BA64E3EG128750
3
T
O
1
D
HMM-985
P
N
NA
NA
NA
1G4AL19E4ED440542
3
S159
O
1
D
158-CEB
P
Y
Y
W
104
1FABP52UXJG185500
6
Q
O
1
D
273-CMS
P
N
NA
NA
NA
1G3AJ54N5L6319366
6
Q
M
1
D
661-CM0
P
NA
N6
NA
NA
1G3WR54TXL0373223
3
T
M
D
15313-T
P
Y
Y
S
108
1GCEK14L5EJ110666
6
Q
M
1
D
795-CYB
P
Y
Y
NA
YEL8
1G3HY54C9LH309730
3
T
M
1
F
764-CMK
P
Y
Y
137
1N4GB22B9LC742425
3
T
M
D
HE-9011
F
NA
NA
NA
NA
F15YKB54013
3
S166
M
D
UN-2888
P
Y
Y
W
138
1GCBS14E3G2140458
3
T
M
1
F
958-CMH
P
N
NA
NA
NA
JG1MS2460LK743586
3
S168
M
1
F
268-DRE
P
Y
Y
M
140
JF1AX9424JG316659
6
Q
M
1
D
712533(ID)
F
NA
NA
NA
NA
3
T
M
2
D
761-CDS
P
N
NA
NA
NA
1B4FK40K6JX292532
3
S171
M
2
D
016-CEA
P
Y
Y
W
116
2B4FK41K2JR589231
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
VtN
Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Confirmed
Odomete
by Decoder
S 173
3
T
M
3
F
LVE-645
P
N
NA
NA
NA
JN1HZ14S2GX141618
NA
S 174
3
T
M
1
D
LPS-022
F
NA
NA
NA
NA
6K92L160868
NA
S 175
3
S173
M
2
F
516-AFW
P
Y
Y
S
134
JN1PS26S4EW633444
92332
S 176
3
S174
M
2
D
CNP-636
P
Y
Y
S
114
256690C110112
No
82386
S 177
6
Q
M
1
D
877-CMT
P
Y
Y
NA
YEL10
1G1LW14T8LE150508
33957
S 178
3
T
M
1
F
143-C4D
P
Y
Y
M
107
JF1AC43B1JC230491
33229
S 179
3
T
M
2
D
98218-N
P
Y
Y
W
119
1GCBS14E5G2149162
72618
S 180
3
T
M
2
F
74547-T
P
Y
Y
S
141
JM2UC2217C0554694
130421
S 181
6
O
M
1
0
179-CYD
P
Y
Y
NA
YEL13
3G4AH54N7LS609966
33836
S 182
3
T
M
1
F
F
NA
NA
NA
NA
NA
S 183
6
O
M
1
D
P
Y
N1
NA
NA
1G1LT54G2LE134260
NA
S 184
3
S182
M
1
F
74685
P
Y
Y
S
115
RA42-061277
95909
S 185
3
T
M
1
F
253-CYE
N
NA
NA
JHMAD5334EC087076
NA
S 186
3
S005
M
1
F
WOQ-249
F
NA
NA
NA
NA
JT2AE82E3G3336773
NA
S 187
3
S025
M
2
0
LVB-147
P
Y
Y
W
146
1G8CT18R5G8134828
74266
S 188
6
Q
O
1
0
2024
F
NA
NA
NA
NA
1FACP52U7LG149271
NA
S 169
3
8005
M
1
D
579-CDY
P
Y
Y
M
109
1FAPP23J6JW186397
32962
S 190
3
S030
M
1
F
CZE-800
N
NA
NA
NA
SJO-3038E96
NA
S 191
3
S043
M
2
D
33325-J
F
NA
NA
NA
NA
1FTBR10C5JUB67141
NA
S 192
6
Q
M
1
D
435-DRD
P
N
NA
NA
NA
1G3NL54U3LM758872
NA
S 193
3
T
M
2
F
33377-J
P
N
NA
NA
NA
1N6ND11S4JC361129
NA
S 194
3
S193
M
2
F
98228 N
P
Y
Y
S
127
JM2UF2111 GO523195
60123
S 195
3
T
M
2
F
VN-1783
P
N
NA
NA
NA
JT4RN50R3G0127582
NA
S 196
3
S043
M
2
D
SMF-573
Y
Y
126
U15SLFE3261
31573
S 197
3
S195
Y
2
F
74030T
P
Y
Y
S
132
RN47-029466
35410
S 198
6
Q
M
1
D
643 -DKS
P
N
NA
NA
NA
1FACP5345LG123282
NA
S 199
6
Q
M
3
D
781-CMH
P
N
NA
NA
NA
1G1FP23T7LL133597
NA
S 200
3
T
Y
1
D
LDH-203
Y
Y
S
122
1X115A6114134
43062
S 201
6
O
O
1
D
422-CMO
P
Y
Y
NA
YEL11
1G3H Y54C9L1816555
51226
S 202
6
O
O
1
D
490-CMT
P
N
NA
NA
NA
1MEPM6047L8619434
NA
S 203
3
T
M
1
0
LVF-557
Y
Y
W
112
1FABP3497GW162154
410
S 204
3
S030
M
2
F
556-EAV
P
Y
Y
M
102
5372057163
119512
S 205
3
T
M
2
F
N7-LAX
Y
Y
S
110
JM2UF1119G0634121
63103
S 206
6
Q
M
1
961-CMK
P
Y
Y
NA
BLU7
2FAPP36X2LB167744
32827
S 207
3
T
Y
1
F
EVT-598
P
Y
Y
S
131
JN1HT14S3DT1038B4
90561
S 208
3
T
M
1
F
396-BNA
P
N
NA
NA
NA
JF2AN53B3JE427176
NA
S 209
3
S208
M
1
F
HZY-065
P
Y
Y
W
128
17A09B7815
102662
S 210
6
Q
M
1
F
738-CYD
P
Y
Y
NA
GRN17
JM1BG2261L0142055
32356
S 211
3
T
Y
1
F
275-AFS
P
Y
Y
S
143
JM1BF2228G0153432
115135
S 212
3
T
M
1
911-CDT
P
Y
Y
S
144
1M07VA7159960
67192
S 213
3
T
M
2
F
UN-2620
N
NA
NA
JM2UF111XG0653535
NA
S 214
3
S213
M
2
F
99716-N
P
Y
Y
W
135
1N6HD11YXJC317900
13007
S 215
6
O
M
1
F
043-CYA
P
Y
Y
NA
ORG18
1N4GB22BXLC764143
14010
-------
Table 4-2 (Continued)
Sample
logger
Seleclion
Driver
Vehicle
Vehicle
tfcense
Vehicle
Oriver
Logger
Distance
Logger
VIN
VIN
Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Confirmed
Odometer
by Decoder
S 216
6
O
O
1
D
466-CYD
P
N
NA
NA
NA
1MECM50U7LGM075S
NA
S 217
6
a
M
1
0
464-CYB
P
Y
Y
MA
YEL12
1G2WJ54T3LF2S3419
23181
S 218
3
T
Y
1
F
006-CYE
P
Y
Y
S
129
SMK-2009776
88373
S 219
3
T
M
1
D
869-CEB
P
Y
Y
S
130
1W27MBK407920
3851 1
SZ20
e
Q
M
1
0
904-CYB
P
Y
Y
MA
BLU3
1MEPM36X2LK631896
3303&
S 221
3
T
M
1
D
526-OKX
P
Y
Y
W
118
1G1AD35P5EJ104025
70033
S 222
3
T
M
2
F
27118P
P
Y
Y
s
117
JM2UF1136K0765114
49467
S 223
3
T
M
1
D
WCG-212
P
N
NA
NA
NA
1Y1SK1943GZ168599
NA
S 224
3
S223
M
1
D
310-BFU
P
Y
Y
W
SN03
1FABP64T8JH140573
56393
S 225
3
T
M
2
D
C&a-CYE
P
N
NA
MA
NA
2P4FH4538LR56O0O1
NA
S 226
3
S225
M
2
D
33392-J
P
Y
Y
W
142
1FTBR10C9JUB9S316
47978
S 227
3
r
V
3
D
215-BND
P
Y
Y
W
SN07
1FABP40A4JF22B320
32750
S 228
3
T
M
2
F
15}03-U
P
N
NA
UA
NA
JT4 VNS3C QJ0020668
NA
S 229
3
6228
M
2
F
LN1B85
P
V
Y
S
125
JM2UC121BE082450Q
32016
S£30
6
Q
M
1
D
636-DKP
P
Y
N1
MA
NA
1GtLTS4G1LY209139
NA
S 231
6
a
M
1
D
714-CMG
F
NA
NA
WA
NA
2G1Vlt54R7U12838e
NA
S 232
3
T
M
1
D
HZC-672
F
NA
NA
MA
NA
JP3BA24K7FU109165
NA
S 233
3
S232
M
1
D
110-DRG
P
Y
Y
W
103
1P3BS4BKXJN143963
76472
S 234
3
T
M
1
D
HMM-993
P
N
NA
NA
NA
1G1AZ3798CR106983
NA
S 235
3
S234
M
1
D
WRS-240
P
Y
Y
W
123
2G2AF51R0H9223691
110928
S 236
6
a
M
1
D
HEIDIS
P
Y
Y
NA
YEL19
1G1LV14T1LE144679
27250
S 237
3
T
M
f
087-CYD
P
N
NA
NA
NA
JT2SV21E3L3419214
NA
S 238
3
S237
M
1
D
591 -BNE
P
Y
Y
W
SN04
1C3CJ41E6JG327477
25200
S 239
3
T
Y
1
D
182-CMN
P
Y
Y
W
SN02
1G1JF77W1GJ206794
102508
S 240
3
T
M
1
F
GOT2SKI
P
Y
Y
M
SN01
JT2ST65L5G7024536
96145
S 241
3
T
M
D
785-AFN
P
Y
Y
M
106
1FM0U15Y1ELA53608
57262
S 242
6
Q
O
1
D
533-DDP
P
N
NA
MA
NA
1FAC P52L) 1L G262150
NA
S 243
6
a
u
1
0
366-CMU
P
Y
Y
MA
YEL11
2G1WN54T9L9201394
13292
S 244
3
T
M
1
D
038-BNG
P
Y
Y
W
133
1FABP55UBJG122178
61692
S 245
3
T
M
1
D
427-AFZ
P
N
NA
NA
NA
183BV41B9CG146685
UA
S 246
3
S245
M
t
D
939-SKH
P
Y
Y
W
124
1FABP52D3JG173298
51306
S 247
3
S031
U
1
D
98-CYL
F
NA
NA
NA
NA
1G3AN69Y1EX334733
MA
N1 V/N not acceptable
N2 No cwnimirticalion with togger
N3 Unable to pick up speed pulses
N5 Error codes
N6 Digital dash
N7 Unaoceplab'.e engine family
NB Unacceptable model year
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logge
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Install
B 001
3
S043
M
2
D
020-171
F
NA
NA
B 002
3
S043
M
2
D
1 BA-274
P
Y
Y
B 003
6
Q
O
1
D
F
NA
NA
B 004
6
Q
O
1
D
NZA-521
P
N
NA
B 005
6
Q
0
1
D
PBB-481
P
Y
Y
B 006
6
Q
M
1
D
NWY-629
F
NA
NA
B 007
3
T
M
1
F
775-AHH
P
Y
Y
B 00B
6
Q
O
1
D
422-AKD
P
Y
Y
B 009
6
Q
1
D
WVB-602
P
N
NA
B 010
3
T
M
1
F
XMG-192
P
Y
Y
B 011
6
Q
M
3
D
VAL-JM
P
Y
Y
B 012
3
T
O
1
0
TNC-883
P
N
NA
B 013
6
Q
M
1
F
YVV-221
P
N
NA
B 014
6
Q
Y
3
D
VTF-944
P
Y
N3
B 015
3
M
1
D
547-AMY
F
NA
NA
B 016
3
B015
M
3
D
WTY-210
P
Y
Y
B 017
a
S005
M
1
F
934-ANC
P
Y
Y
B 018
3
B012
O
1
0
TYC-278
P
N
NA
B 019
3
T
M
1
D
YtW-084
P
Y
Y
B 020
3
B012
O
1
D
SDC-293
P
N
NA
B 021
3
B012
o
1
D
72S-ADW
P
N
NA
B 022
3
B012
o
1
D
BBO-267B
P
N
NA
B 023
3
T
M
2
D
348-204
P
N
NA
B 024
3
B012
O
1
0
P
N
NA
B 025
6
Q
M
1
D
F
NA
NA
B 026
6
Q
M
1
D
ZKF-718
F
NA
NA
B 027
3
T
M
2
0
12739B-M
P
N
NA
B 028
6
Q
M
2
D
297412-M
P
Y
N5
B 029
3
S023
M
2
0
887-123
F
NA
NA
B 031
3
B012
O
1
F
KCO-1718
P
N
NA
B 032
3
B012
O
1
0
33997
P
N
NA
B 033
3
B023
M
2
D
596-895
F
NA
NA
B 034
3
B012
O
1
D
RYH-602
F
NA
NA
B 035
3
B012
O
1
D
VAN-545
P
N
NA
8 036
3
T
M
2
D
437265-M
P
N
NA
B 037
3
S023
M
2
D
WMC-0051
P
Y
Y
B 038
3
M
1
D
WVB-977
P
N
NA
B 039
3
B012
O
1
D
RNT-659
P
Y
Y
B 040
PYX-101
B 041
3
S023
M
2
0
AAG-32T
F
NA
NA
B 042
3
O
1
D
HHS-880
P
Y
Y
B 043
1
D
299-AFJ
P
N
B 044
6
Q
O
1
D
00709
P
Y
Y
Distance Logger VtN
Sensor Number
MA
NA
1FTDF15Y4HNA67137
W
105
1FTCR10A1MUC48371
NA
NA
NA
NA
2G4WD14W7K1435099
NA
RED2
3B3XA4633MT034156
NA
NA
1B3BA76J4KF443588
M
104
JHMBB7232GC026142
NA
YEL12
1G1LT64W8KY218187
NA
NA
1G2NE14D3KC752033
W
111
YV1AX885XJ1793960
NA
RED1
tB3XG24K5KG 180041
NA
NA
1FABP57U4JA223282
NA
NA
1YVGD22B9M5123943
NA
NA
1FABP44A6KF128781
NA
NA
1FAPP9194KT137977
W
145
1G1FP21S3KL104669
M
113
YS3AL75L8M7006092
NA
NA
1G3HN54C6M1812733
W
134
1G1FP21S1JL193166
NA
NA
1G4 BRB1Y3 K A404971
NA
NA
1G6CD13B2M4277240
NA
NA
1G3CW53L4M4318426
NA
NA
1B7FD04H6GS06728D
NA
NA
NA
NA
NA
NA
2G1WN54TCL9311377
NA
NA
2B6HB23TXHK266154
NA
NA
2B4FK25K7K R269211
NA
NA
1FTEF15N6JNA59850
NA
NA
1G3AJ51W6KG324243
NA
NA
2MECM74F5MX661232
NA
NA
1GCCT14B5E2173384
NA
NA
3B3BK46D3KT993747
NA
NA
2MEBM75F6KX675910
NA
NA
1GNCT18Z5M8146406
W
107
2P4FH4t3XJR662190
NA
NA
1 FABP52U8KA158932
W
142
1P3BK46D2KC469869
2FABP74F8KX183167
NA
NA
2P4FH51G4GR751156
S
135
1G6AD6983D9266654
NA
NA
1FACP57U3MA155663
NA
YEL11
1G2HZ54C2KW215880
VIN Final
Conlirmed Odomeler
by Decoder
NA
21179
NA
NA
6710
NA
46819
35220
NA
761 GO
61720
NA
NA
NA
71404
15011
NA
198B3
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
42865
NA
24327
NA
71701
29495
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logge
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Install
B 045
6
Q
M
1
F
YTC-431
P
N
NA
B 046
e
Q
M
1
D
REC-404
P
N
NA
B 047
6
Q
M
1
D
RPW-065
P
Y
N1
B 048
3
B03G
M
F
TAP-B21
P
Y
Y
B 049
3
T
M
1
F
F
NA
NA
B 050
3
S023
M
D
81658
P
Y
Y
B 051
3
T
M
1
D
MJK-JR
F
NA
NA
B 052
3
B049
Y
1
F
YRE-029
P
Y
Y
B 053
6
Q
M
1
D
JHU-0593
P
Y
Y
B 054
3
M
286847-M
P
N
NA
B 055
6
Q
Y
1
F
484-AJW
P
Y
Y
B 056
6
Q
M
1
D
P
N
NA
B 05?
3
T
M
1
D
5T05567
F
NA
NA
B 056
3
M
1
D
ZPH-57E
P
Y
Y
B 059
3
B057
M
1
D
0)999
P
Y
Y
B 060
e
Q
M
1
F
WXN-300
F
NA
NA
B 061
3
S023
M
D
AAH-04J
P
N
NA
B 062
3
M
1
0
PBN-705
P
Y
Y
B 063
3
S023
M
D
810954
P
N
NA
B 064
3
S023
M
D
264-358
P
N
NA
B 065
6
Q
M
1
0
RYE-253
P
Y
Y
B 066
3
BO 49
M
1
F
YSZ-145
P
Y
Y
B 067
3
T
M
1
0
YRZ-855
P
N
NA
B 068
3
S067
1
D
653-ABT
F
NA
NA
B 069
3
B067
M
1
D
P
N
NA
B 070
3
B067
M
1
0
839-ACX
P
N
NA
B 071
3
B067
M
1
D
YTZ-547
P
Y
N
B 072
3
T
M
D
982-493
P
Y
Y
B 073
6
Q
M
1
D
XCV-403
P
N
NA
B 074
6
Q
M
1
D
YSZ-538
P
Y
Y
B 075
3
T
P
N
NA
B 076
6
O
M
1
D
YPX-636
P
Y
Y
B 077
3
T
M
1
D
VXA-316
F
NA
NA
B 076
3
B071
M
1
0
VWH-608
P
N
NA
B 079
3
S023
M
D
327-664
P
Y
Y
B 080
6
O
M
1
D
WTX-579
P
N
NA
B 081
3
M
1
D
WVF-128
P
Y
Y
B 082
6
0
M
1
D
YSL-581
F
NA
NA
B 083
3
T
M
1
D
RBM-985
P
N
NA
B 084
3
B083
M
1
D
SMZ-544
P
Y
Y
B 085
3
M
1
D
XXR-316
P
N
NA
B 086
3
B071
M
1
D
WXK-557
P
Y
Y
B 087
6
Q
O
1
D
KCO-5595
P
N
NA
Dlslance
Sensor
Logger VIN
Number
VIN Final
Confirmed Odomeler
by Decoder
NA
NA
JM1BG2262M0228959
NA
NA
NA
1 FABP50U8KA1779S7
NA
NA
NA
2G4WB14LXM1861528
NA
W
108
1N6SD11SBMC323930
18903
NA
NA
NA
S
116
1GCEG25HXC7125742
78563
NA
NA
1FABP44E1KF11499B
NA
W
140
1HGCB7661MA049561
25555
NA
YEL9
1G3NT14DXKM255041
25661
NA
NA
1GNCT18R6K0155353
NA
NA
GRN16
JM1GD2224L1810193
34399
NA
NA
NA
NA
NA
INTERCHANGEABLE
NA
W
117
1G2AB2705E7332521
81575
W
123
1G1JF11W2K7180957
29442
NA
NA
1N4GB22S8KC733973
NA
NA
NA
1FMCU14T6GUB88128
NA
W
134
1FAPP14J2MW269094
4365
NA
NA
1FTCR14A6GPA92258
NA
NA
NA
2GCCC14D5C1177157
NA
NA
BLU5
1FAPP6046KH126728
34728
S
119
1YVGD31B5M513B706
9633
NA
NA
1G3EV13L3MU306160
NA
NA
NA
1G2FS23EXML205941
NA
NA
NA
NA
NA
NA
JN1MS34PBMW003819
NA
NA
NA
1FAPP1283MW223855
NA
W
139
1FTCR10A1KUA84004
33205
NA
NA
1MEBM55U2KA601525
NA
NA
BLLI4
1MEPM6042LH679380
15675
NA
NA
NA
NA
BLU4
1FACP57U7MA175592
1BB55
NA
NA
1G1AE27P1E7260581
NA
NA
NA
1G1JC1111K7158004
NA
M
141
1FTBR10T6GUB61390
553
NA
NA
1LNBM83FXK Y727917
NA
W
129
1G1JC1113KJ171405
46260
NA
NA
1G1AW81W6K6214925
NA
NA
NA
1B3BA64 E2EG275318
NA
S
128
1G2AB2707E7250564
B0424
NA
NA
1FABP29U1GG278824
NA
W
114
1FAPP36X9KK201666
601 16
NA
NA
1G3HN54C6KW3574B0
NA
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
VIN Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Confirmed Odomete
by Decoder
B 068
3
T
M
2
D
01350
P
Y
Y
W
127
1GCDC14Z8KZ211565
37048
B 089
6
Q
O
1
D
TLV-278
P
N
NA
NA
NA
1G2HX54C4L1219871
NA
B 090
6
Q
Y
1
D
WDG-803
F
NA
NA
NA
NA
1G3WH14TXMD308045
NA
B 091
6
O
M
1
D
PE00019
F
NA
NA
NA
NA
1G4HP54C1MH431160
NA
B 092
6
O
O
1
D
YTG-991
F
NA
NA
NA
NA
1G3HN54C7MH335477
NA
B 093
3
O
1
D
SFT-453
F
NA
NA
NA
NA
1G1JC5116K7194911
NA
B 094
6
Q
o
1
F
RYR-967
P
N
NA
NA
NA
1N4GB21S8KC733988
NA
B 09S
3
T
M
D
765-713
F
NA
NA
NA
NA
1FTBR10S3EUB42008
NA
B 096
3
B095
M
D
AAF-390
F
NA
NA
NA
NA
NA
B 097
3
M
1
0
YSZ-722
F
NA
NA
NA
NA
2G1WL54T4M1118073
NA
B 098
3
B095
O
D
683-450
P
Y
Y
W
112
1GTBS14R2G2533452
12968
B 099
6
Q
M
1
D
TJD-509
P
Y
Y
NA
YEL10
1G3HY14C3KW311671
37534
B 100
3
T
O
1
D
SAJ-329
F
NA
NA
NA
NA
7E35F526110
NA
B 101
6
Q
M
1
D
025-AKY
P
Y
Y
NA
YEL13
1G3CW54C8K4300538
35425
B 102
3
T
M
1
F
TBN-374
P
N
NA
NA
NA
1N4GB21S5KC782419
NA
B 103
3
T
M
D
469-355
F
NA
NA
NA
NA
CE239B869962
No NA
B 104
3
B051
M
1
D
RJW-112
P
Y
Y
110
1G3A J1137HD381642
50422
B 105
3
T
M
F
384-613
P
Y
Y
S
126
JT4RN50R7G0173173
62147
B 106
6
Q
M
1
D
TYK-159
P
N
NA
NA
NA
1G1LT53T4MY110010
NA
B 107
6
Q
M
1
D
ZJR-240
F
NA
NA
NA
NA
1G2WH54T9M F200743
NA
B 108
6
Q
O
1
D
PCB-113
P
N
NA
NA
NA
1G4AL51N1KT476385
NA
B 109
6
Q
O
1
0
PXJ-491
P
N
NA
NA
NA
1G2HX54C9KW206714
NA
B 110
3
B100
M
1
D
NWY-631
P
Y
Y
M
109
1H24KA4409308
6315
B 111
3
T
M
1
F
1A34725
F
NA
NA
NA
NA
INTERCHANGEABLE
NA
B 112
6
Q
O
1
D
KC00531
P
Y
N
NA
NA
1G4NC54N0KM036312
NA
B 113
6
Q
M
1
F
ZEJ-510
P
Y
Y
NA
ORG18
1N4GB2253KC735159
103126
B 114
3
M
1
D
YVY-911
P '
N
NA
NA
NA
1G3AM19E4EG316495
NA
B 115
3
B114
M
1
D
WFF-409
P
N
NA
NA
NA
1G1AZ3791DB110316
NA
B 116
3
B114
M
1
D
RX00253
P
N
NA
NA
NA
1G4AL19X3GT414184
NA
B 117
3
B114
M
1
D
PEB-223
P
Y
Y
W
106
1G1JC8118H7116530
81318
B 118
6
Q
M
1
D
WPC-937
P
N
NA
NA
NA
1G1JC111XKJ195751
NA
B 119
6
Q
M
1
D
ZJL-454
P
Y
Y
NA
YEL8
1G1JF14T4M7133492
26523
B 120
3
T
M
D
431250 M
P
N
NA
NA
NA
2B4GK25KXMR130052
NA
B 121
3
B111
M
1
F
PMA-712
F
NA
NA
NA
NA
YV1AX884XH1244342
NA
B 122
3
M
1
F
XLV-069
F
NA
NA
NA
NA
JT2AL22G2B2222699
NA
B 123
3
B111
M
1
F
ITWORKS
P
Y
Y
S
144
TE72-509680
No 100399
B 124
3
T
O
1
D
PLR-789
P
Y
Y
W
118
1B3BP48D5KN589752
17117
B 125
3
T
M
1
D
YVE-420
F
NA
NA
NA
NA
1FAPP36X0MK158001
NA
B 126
3
B125
M
1
D
P
N
NA
NA
NA
NA
B 127
3
B125
M
1
D
YTG-569
P
N
NA
NA
NA
1G1JC54G6MJ194851
NA
B 128
3
B120
M
2
0
295251-M
P
N
NA
NA
NA
1FMCU14T5KUA42666
NA
B 129
3
M
2
F
134413
F
NA
NA
NA
NA
JT4RN67PXH5048277
NA
B 130
3
B125
M
1
D
WTN-186
P
Y
Y
W
137
1FABP44E2KF117098
45568
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
VIN Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Confirmed Odomete
by Decoder
B 131
3
B103
M
2
0
453 026
F
NA
NA
NA
NA
F25HEZ04095
NA
B 132
3
T
M
1
D
XVZ-941
P
N
NA
NA
NA
1G1GZ37G2GR110555
NA
B 133
3
B132
M
1
D
WSV-240
P
N
NA
NA
NA
1G1AZ3799DB142009
NA
B 134
3
B132
M
1
D
RYS-953
P
N
NA
NA
NA
1G3AK47Y7EM450276
NA
B 135
3
B132
M
1
D
XYZ-232
P
N
NA
NA
NA
1G1TB68C9FA154310
NA
B 136
6
Q
O
1
D
RKB-107
P
N
NA
NA
NA
1LNCM81W8MY714626
NA
B 137
3
B125
M
1
D
TGZ-942
F
NA
NA
NA
NA
NA
B 138
3
T
M
D
235-550
P
N
NA
NA
NA
CLN14A8241066
NA
B 139
3
B132
0
1
0
TVR-667
P
Y
Y
W
143
1G4AL1937F6463396
46927
B 140
6
Q
O
1
D
WTL-100
P
Y
Y
NA
BLU7
1MEPM36X2KK620959
43545
B 141
3
T
M
1
D
NET-237
F
NA
NA
NA
NA
1B3BZ48C6GD222454
NA
B 142
6
Q
M
1
D
VAM-280
F
NA
NA
NA
NA
1FABP52U2KA159753
NA
B 143
3
T
M
1
F
270AMZ
P
Y
Y
S
121
JT2RA44C6B0001018
83805
B 144
6
Q
0
1
F
TJT-292
P
Y
Y
NA
BLU15
JE3CU36X6KU035123
35231
B 145
6
Q
M
1
D
TSB-995
P
Y
Y
NA
YEL9
1G3NT14D2KM235737
16395
B 146
3
T
M
1
F
093-ACR
P
N
NA
NA
NA
JT2ST65C5H7106008
NA
B 147
3
S030
M
1
F
ZFB-153
P
N
NA
NA
NA
1G6DW51Y0J9713286
NA
B 148
3
B142
M
1
D
XPP-662
F
NA
NA
NA
NA
1FABP40A1HF150461
NA
B 149
3
B149
M
1
F
VKV-489
F
NA
NA
NA
NA
JHMSM5429BC042513
NA
B 150
3
B142
M
1
D
WPK-B51
P
Y
Y
W
122
1 B3YA44K6JG448546
46752
B 151
3
M
D
340220-M
F
NA
NA
NA
NA
J9F17NH084652
NA
B 152
3
7
M
1
F
SXH-363
P
N
NA
NA
NA
1HGCA6281KA019512
NA
B 153
3
B138
M
D
AAG-51V
P
Y
Y
146
1GNCT18Z3K0152538
52626
B 154
6
Q
1
D
YLL-849
P
Y
N
NA
NA
1G3WH54T6MD326808
NA
B 155
3
T
M
1
D
ZEP-6S7
NA
NA
NA
1FABP28M2F1553800
No NA
B 156
3
T
M
D
998-474
P
Y
Y
W
103
1GCFC24H1KE115676
43204
B 157
6
Q
1
D
166-AHD
P
N
NA
NA
NA
1G3HN54C9MH308961
NA
B 158
3
T
M
1
0
SWP-680
F
NA
NA
NA
NA
1G3AJ51W8KG342307
NA
B 159
- 6
Q
O
1
D
19941
P
Y
Y
NA
YEL11
1G2HX54CXKW259695
27252
B 160
3
B120
M
D
431478-M
P
N
NA
NA
NA
2B4GK2536MR172629
NA
B 161
3
B120
M
D
096-846
P
N
NA
NA
NA
1GTDK14K6KE527575
NA
B 162
3
M
1
0
TYV-037
P
N
NA
NA
NA
1MEPM6040MH619227
NA
B 163
3
B146
M
1
F
VRH757
P
Y
Y
W
124
JHMCA53B6JC052224
88404
B 164
3
T
O
1
D
RYB-829
P
Y
Y
W
130
1B3BD31D6FG241191
58540
B 165
6
Q
M
D
414-713
P
Y
Y
NA
ORG1
2B4GK55R7MR201092
13888
B 166
3
T
O
1
D
SVT-205
P
N
NA
NA
NA
1P3BP36DXHF290238
NA
B 167
6
Q
M
1
F
YSV-199
P
Y
Y
NA
GRN17
1YVGD22B0M5125967
15186
B 168
3
B166
O
1
F
RKY-116
F
NA
NA
NA
NA
1C3BF66P5FX553570
NA
B 169
3
B166
O
1
0
RYC-555
P
Y
Y
W
125
1G4XB69R0FW4 54358
78872
B 170
3
S110
M
1
F
WVX-385
P
N
NA
NA
NA
WVWEA016XJW311553
NA
B 171
3
M
2
D
321457-M
P
N
NA
NA
NA
1FMDU34X1MUC9540B
NA
B 172
3
S025
Y
2
F
978-345
P
N
NA
NA
NA
JT2RN81R5K5006631
NA
B 173
6
Q
M
1
D
P
N
NA
NA
NA
NA
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
VIN
Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Conlirmed
Odomete
by Decoder
B 174
3
M
2
D
018-115
P
N
NA
NA
NA
CCU140B158579
NA
B 175
3
T
M
1
F
ZHB-646
P
N
NA
NA
NA
1HGED3659JA02G883
NA
B 176
3
M
1
D
TSR-546
P
N
NA
NA
NA
1P3BM44C3ED293272
IMA
B 177
3
S110
M
1
F
VLF-644
P
N
NA
NA
NA
YV1FX8848H2186523
NA
B 178
3
B153
O
1
F
WVT-348
P
Y
Y
W
131
JH4DA3458KS013350
20744
B 179
3
T
M
1
D
WVT-955
F
IMA
NA
NA
NA
1G1BN51E6KA131084
NA
B 180
a
B179
M
1
D
610-ANB
F
NA
NA
NA
NA
1G3WS14W8K D345586
NA
B 181
3
B179
M
1
D
XMF-070
P
N
NA
NA
NA
1Y1SK5149JZ025621
NA
B 182
3
M
1
D
P
N
NA
NA
NA
NA
B 183
3
B179
M
1
D
WVA-200
P
N
NA
NA
NA
1G2NE14U6KC713016
NA
B 184
3
B179
M
1
D
TCM-132
P
Y
Y
W
133
1G4AH51R1HT438622
54376
B 185
3
T
M
D
631-875
P
N
NA
NA
NA
1GCCW80H7ER108666
NA
B 186
3
M
1
D
SEN-341
F
NA
NA
NA
NA
1G1JC1117H7159408
NA
B 187
6
Q
O
1
0
WVD-342
P
N
NA
NA
NA
1G4AL51N1KT468805
NA
B 188
6
Q
M
1
D
WRS-182
P
N
NA
NA
NA
1G3AJ51N6KG330762
NA
B 189
6
Q
1
F
XXR-511
P
N
NA
NA
NA
JN1HJ01P9LT376777
NA
B 190
3
T
Y
1
F
YRD-994
P
Y
Y
W
120
JT2EL34B2M0027569
15264
B 191
3
B120
M
D
367-625
P
Y
Y
W
136
1GCDC14H3ME100424
12119
B 192
6
Q
M
1
D
YWJ-196
F
NA
NA
NA
NA
1G1LV13T6MY139222
NA
B 193
3
B185
M
D
981-043
P
Y
Y
W
SN09
1GCDC14ZSKE135013
45189
B 194
6
Q
1
0
PLY-781
P
N
NA
NA
NA
1G2NE54U6MC527177
NA
B 195
3
T
Y
1
F
P
Y
Y
W
SN06
JHMBA7431GC066566
112991
B 196
6
Q
M
1
D
YSV-377
P
Y
Y
NA
YEL11
2G1WN14T9L9302959
15423
B 197
6
Q
M
0
TXD-453
F
NA
NA
NA
NA
1G1JF11WXK7174673
NA
B 198
6
Q
M
1
D
862-AMD
P
N
NA
NA
NA
1G4CW54C7K1639319
NA
B 199
6
Q
M
1
D
064-ACS
F
NA
NA
NA
NA
1G2WJ54T1M F245132
NA
B 200
3
T
O
1
D
XXR-189
P
N
NA
NA
NA
1G4JS 5116KJ409647
NA
B 201
6
Q
M
1
D
WTP-491
P
Y
N5
NA
NA
1G3HY14C2KW368282
NA
B 202
3
M
1
F
WVT-163
F
NA
NA
NA
NA
JM1BD2310E0760014
NA
B 203
3
M
1
D
VHE-395
F
NA
NA
NA
NA
1G3AM19EXDD420084
NA
B 204
3
M
1
D
XKE-337
P
N
NA
NA
NA
1G4NM27 H7FM478057
NA
B 205
3
M
1
D
RRH-857
NA
NA
NA
1G3NT 54L4 JM259390
NA
B 206
3
T
O
1
D
12820
P
Y
Y
W
SN05
2MEBM75F1KX665883
41294
B 207
3
S076
M
1
D
XXD-452
P
Y
Y
W
SN04
JG1MR2152HK746455
33132
B 208
3
Y
1
F
928-AEE
P
Y
Y
S
SN02
JM1FC3312G0121078
111384
B 209
3
T
M
1
0
VDP-360
NA
NA
NA
NA
B 210
3
T
O
1
F
123-AEC
P
Y
Y
M
SN01
JF1AC42B2KC210379
33445
B 211
6
Q
M
1
D
YRA-867
P
N
NA
NA
NA
1MEPM60T2MH631032
NA
B 212
6
Q
0
PAX-674
P
N
NA
NA
NA
1B3XG24K7KG107124
NA
B 213
6
Q
1
D
YYC-671
P
N
NA
NA
NA
1G3AJ51 N2KG325381
NA
B 214
6
Q
M
D
418626-M
P
Y
N5
NA
NA
1P4GH54R0MX579786
NA
B 215
3
T
M
1
D
VFB-404
F
NA
NA
NA
NA
6D47S8E661372
NA
B 216
6
Q
1
D
SFX-142
P
N
NA
NA
NA
3G4AH54NXMS603113
NA
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
VIN Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Confirmed Odomete
by Decoder
B 217
3
B215
M
1
D
ZHV-260
P
Y
Y
M
105
1FABP1323DT120604
63726
B 210
6
Q
1
D
YTG-870
F
NA
NA
NA
NA
1G2JC14K9M7585174
NA
B 21 9
6
Q
1
D
YVE-100
P
N
NA
NA
NA
2G4WD54LXM1855715
NA
B 220
6
Q
M
1
D
ZHB-267
F
NA
NA
NA
NA
1B3XA4630MF584988
NA
B 221
6
Q
O
1
D
SLR-603
P
N
NA
NA
NA
1G4HR54C8MH401146
NA
B 222
6
Q
1
D
WVT-103
P
N
NA
NA
NA
1G1JC111XKJ215111
NA
B223
6
Q
M
1
D
XKD-737
F
NA
NA
NA
NA
1G4AH51N7KG40S800
NA
B 224
6
Q
1
D
XMY-483
P
N
NA
NA
NA
1G2FS23T0ML212461
NA
B 225
3
T
M
1
F
ZFK-055
P
N
NA
NA
NA
YV1FX884SG2051945
NA
B 226
6
Q
M
1
D
YNT-239
F
NA
NA
NA
NA
2G1WL54T1M9113477
NA
B 227
6
Q
1
D
970-ALM
P
N
NA
NA
NA
NA
B 228
3
B175
M
1
F
WYN-470
P
N
NA
NA
NA
JF1AC45B7HC202556
NA
B 229
3
B175
M
1
F
VRV-952
P
N
NA
NA
NA
1HGED3553JA033488
NA
B 230
6
Q
M
1
D
PZY-943
P
N
NA
NA
NA
1G3AJ51W7KG320122
NA
B 231
3
B225
M
1
F
YJR-728
F
NA
NA
NA
NA
JT2AE86S9E0049002
NA
B 232
6
Q
M
1
D
WWN-287
F
NA
NA
NA
NA
1G3NF54N4KM270366
NA
B 233
3
B175
M
1
F
XZC-169
P
Y
Y
S
104
JN1GB21SXKU537381
41170
B 234
3
M
1
D
656-ADW
P
Y
N
NA
NA
1G4EZ13L2M4404077
NA
B 235
6
Q
M
1
D
BBO-5587
P
N
NA
NA
NA
1G4NV54N1MM259987
NA
B 236
3
T
M
1
D
VSX-591
P
Y
Y
108
1W19J9B567918
30066
B 237
3
B175
M
1
F
VNT-707
P
N
NA
NA
NA
JM1FC3314J0610704
NA
B 238
6
Q
Y
1
0
ZDG-037
P
Y
Y
NA
RED2
1B3XG24K4KG153848
34873
B 239
3
T
O
1
D
ALO-2624
P
N
NA
NA
NA
2FABP74F1KX108679
NA
B 240
3
M
1
F
WTN-515
P
N
NA
NA
NA
JNl HS34P7KW012610
NA
B 241
3
B239
O
1
D
898-AHN
P
Y
Y
W
145
1MEBM55UXKA633641
52045
B 242
3
T
M
1
D
TRG-457
P
N
NA
NA
NA
1G2FS87 S9 FN215032
NA
B 243
3
B175
M
1
F
XMT-325
P
Y
Y
M
113
YV1AX8852H1723847
66619
B 244
3
B225
M
1
F
PJZ-774
F
NA
NA
NA
NA
KMHLF21J7HU186053
NA
B 245
3
B225
M
1
F
YWJ-392
F
NA
NA
NA
NA
KMHLD11J6HU043837
NA
B 246
3
T
M
D
271641-M
F
NA
NA
NA
NA
1GNCT18Z0KO165229
NA
B 247
3
B246
M
D
VVX-502
P
Y
Y
W
107
2B4FK5131JR676116
51090
B 248
6
Q
1
D
TMV-239
P
N
NA
NA
NA
1FACP50U4LA153137
NA
B 249
6
Q
M
1
D
WVP-840
F
NA
NA
NA
NA
1 FABP54Y4KA190602
NA
B 250
3
M
1
D
ZHZ-332
F
NA
NA
NA
NA
1FABP132XDT137948
NA
B 251
3
B200
O
1
D
00623
P
N
NA
NA
NA
1LNBM81F7JY847577
NA
B 252
3
B225
M
1
F
835-AMM
P
N
NA
NA
NA
1HGED3546KA028309
NA
B 253
3
T
M
F
212-728
F
NA
NA
NA
NA
JT4RN50R2H5105660
NA
B 254
6
Q
O
1
D
TCG-797
P
N
NA
NA
NA
1MEPM6040KH624702
NA
B 255
3
B200
M
1
D
HC00181
P
Y
Y
W
140
1C3BC59KXHF330954
38481
B 256
3
M
t
D
XME-131
P
Y
Y
M
111
2FABP38S0HB102197
60767
B 257
6
Q
M
1
D
F
NA
NA
NA
NA
NA
B 258
6
Q
M
1
F
YSZ-618
P
N
NA
NA
NA
JN1HJ01P4LT430681
NA
B 259
3
B242
M
1
D
VHX-696
P
N
NA
NA
NA
1G1AW19R9D6815769
NA
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
VIN Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Conlirmed Odometer
by Decoder
B 260
6
Q
M
1
D
WRL-765
F
NA
NA
NA
NA
1LNLM9845KY673685
NA
B 261
3
T
M
1
F
YTZ-102
P
N
NA
NA
NA
JT8UF11E3M0066028
NA
B 262
3
B225
M
1
F
WHS
F
NA
NA
NA
NA
YV1GA694XJ0054177
NA
B 263
3
B387
M
1
F
ZHT-936
P
Y
Y
M
103
JHMBA3144HC019202
95996
B 264
3
B242
M
1
D
VXF-522
F
NA
NA
NA
NA
2G2AF27X1E1219504
NA
B 265
3
B242
M
1
D
ZHG-206
P
Y
Y
116
1FABP30U2GA193321
97951
B 266
6
Q
1
D
SDH-424
P
Y
N1
NA
NA
2G4WD54L4M1844158
NA
B 267
3
B253
M
1
D
02380
F
NA
NA
NA
NA
0G87F159842
NA
B 268
3
B242
M
1
0
TAJ-192
P
Y
Y
S
119
1G1AW35K5CR225618
73012
B 269
3
M
0
218-807
P
Y
Y
S
SN03
1B7HD14T7HS461522
111838
B 270
6
Q
M
1
D
YVV-194
P
N
NA
NA
NA
1G4HR54C3MH445510
NA
B 271
3
T
M
1
F
WTK-331
P
N
NA
NA
NA
JN1GB21S0KU532349
NA
B 272
3
B271
M
1
F
SCE-990
F
NA
NA
NA
NA
YV1AX4951C1381436
NA
B 273
3
S196
M
D
989-465
P
Y
Y
W
139
1FTCR11T1JUE26260
27876
B 274
3
M
1
F
YTJ-058
P
N
NA
NA
NA
2T1AE94A7MC084133
NA
B 275
3
S110
M
1
F
VXB-579
P
Y
Y
S
125
JT2AL31G7E0225492
72290
B 276
3
B271
M
1
F
SAZ-279
P
N
NA
NA
NA
KMHLF31J8HU117832
NA
B 277
3
S096
M
1
D
SRJ-807
F
NA
NA
NA
NA
1B3BK46K3KC422625
NA
B 278
3
T
M
1
F
VXN-311
F
NA
NA
NA
NA
JN1HU11S0GT124004
NA
B 279
3
B278
M
1
F
VLD-971
P
N
NA
NA
NA
WBADK8301D9206190
NA
B 280
3
B278
M
1
F
RJP-749
P
Y
Y
M
132
YS3AG43S3D1020742
93409
B 281
6
Q
M
1
D
ZKV-901
P
Y
Y
NA
BLU6
2FAPP35X1MB126332
5372
B 282
6
Q
M
1
D
WPA-613
P
Y
Y
NA
RED1
1FABP55UXKA153310
35945
B 283
3
T
M
1
D
VSY-667
F
NA
NA
NA
NA
2XMJP557XJA020762
NA
B 284
3
B261
M
1
F
PVX-136
P
N
NA
NA
NA
YV1FX8848K2314251
NA
B 285
3
B271
M
1
F
PVM-139
F
NA
NA
NA
NA
JT2MX62E1B0022724
NA
B 286
3
B103
M
D
10151
P
Y
Y
S
SN07
CRN1498264964
13040
B 287
3
T
M
1
D
ZGK-247
P
Y
Y
S
135
1Z37J9B412061
50242
B 288
6
Q
M
1
D
WXE-726
P
N
NA
NA
NA
1FAPP6045KH163835
NA
B 289
6
Q
M
1
F
90KMAXX
F
NA
NA
NA
NA
JN1HJ01P9KT240079
NA
B 290
6
Q
O
1
D
54417
P
N
NA
NA
NA
1G4HP54C1MH424337
NA
B 291
6
Q
M
1
D
WTK-513
P
Y
Y
NA
BLU5
1FAPP36XXKK176728
46185
B 292
3
T
M
1
D
RBK-888
P
N
NA
NA
NA
1G3BY69Y9FY342666
NA
B 293
3
B261
M
1
F
LHS0090
P
Y
Y
M
127
VF3DA2128KS506151
35297
B 294
6
Q
M
1
D
WVP-659
F
NA
NA
NA
NA
1MEPM6041KH604264
NA
B 295
6
Q
O
1
D
TKB-571
P
N
NA
NA
NA
1G3AJ51N9KG340220
NA
B 296
6
Q
M
1
D
XDT-029
P
N
NA
NA
NA
1LNBM81FOKY640370
NA
B 297
3
B292
M
1
D
NCT-340
P
Y
Y
W
129
1MEBP9237GH654018
53335
B 298
3
B271
M
1
F
YHM-975
P
N
NA
NA
NA
1G1LW14T8LY125235
NA
B 299
6
Q
M
1
D
WSA-922
F
NA
NA
NA
NA
1FAPP36X6KK153561
NA
B 300
3
T
M
D
162280-M
F
NA
NA
NA
NA
CGU157U196422
NA
B 301
3
S122
M
1
F
RTJ-407
P
Y
Y
M
138
JHMST5434DS016244
92925
B 302
6
Q
M
1
D
XXB-351
P
Y
Y
NA
YEL13
1G1LT54T2LY160319
33703
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
VIN
Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Confirmed Odomete
by Decoder
8 303
3
M
1
D
SKK-093
P
N
NA
NA
NA
1G1JC5111H71700B6
NA
B 304
3
M
1
D
YEX-849
P
N
NA
NA
NA
1G2HZ54C9HW322790
NA
B 30S
3
B300
M
0
659-744
F
NA
NA
NA
NA
CCL449A162789
NA
B 306
3
B283
M
1
D
NNC-284
F
NA
NA
NA
NA
1MEBP6545DW644243
NA
B 307
6
Q
Y
1
D
WXL-703
F
NA
NA
NA
NA
1G1LV14W6KY180794
NA
B308
3
B283
M
1
D
XKM-260
P
N
NA
NA
NA
1B3BD49D2FF306621
NA
B 309
3
T
M
1
D
SPD-889
P
N
NA
NA
NA
1FAPP2899JW209976
NA
B 310
3
B283
M
1
D
XMF-329
P
N
NA
NA
NA
1C3BH58E4HN480772
NA
B 311
3
B253
M
F
248-767
P
N
NA
NA
NA
1N6ND11Y2GC448136
NA
B 312
3
M
D
208-009
P
N
NA
NA
NA
F10GEZ01518
NA
B 31 3
3
B283
M
1
0
ZHG-122
F
NA
NA
NA
NA
1G4AJ69A6EH621456
NA
B 314
3
B283
M
1
D
SMT-854
P
Y
Y
W
110
1G3CX69B9G4332506
55621
B 31 5
3
T
M
0
340-030
P
Y
Y
W
141
1FTEX15H2GKA9415S
37404
B 316
6
Q
O
1
D
TTJ-779
P
N
NA
NA
NA
1G3AJ51N7KG332410
NA
B 317
3
T
O
1
D
TDD-547
P
Y
Y
M
124
8G87H169118
89019
B 318
3
S130
M
1
D
TSD-727
P
N
NA
NA
NA
1G2BL81Y5JA200424
NA
B 319
3
S130
M
1
D
WRC-593
P
Y
Y
W
114
1FABP259XHW116408
77229
B 320
6
O
M
1
D
SYX-874
F
NA
NA
NA
NA
1MEBM55U9KA632819
NA
B 321
3
T
M
1
F
XYL-899
P
N
NA
NA
NA
YV1AX8855G1681639
NA
B 322
3
T
M
1
F
092-AKW
P
N
NA
NA
NA
JN1PB12S4FU631495
NA
B 323
6
Q
M
1
D
PPF-149
P
N
NA
NA
NA
1FABP55U1KA106599
NA
B 324
S
O
M
1
D
YEJ-185
P
Y
Y
NA
BLU4
1LNBM81F0KY736595
46955
B 325
3
T
M
F
007-933
P
Y
Y
S
128
JT4RN50R7H0325017
57739
B 326
6
Q
O
1
D
ZION
F
NA
NA
NA
NA
1C3XY56R3LD910804
NA
B 327
6
Q
O
1
D
256-AG V
P
N
NA
NA
NA
1P3XA5630MF541806
NA
B 328
6
Q
M
1
0
TKC-237
F
NA
NA
NA
NA
2MEPM36X5KB608556
NA
B 329
3
T
o
1
0
SZG-438
P
Y
Y
W
106
1G1GZ11G2HP118178
25410
B 330
6
O
0
1
0
PBZ-850
P
N
NA
NA
NA
1FABP52U4KA146986
NA
B 331
6
Q
o
1
D
REP-122
P
N
NA
NA
NA
1G3HN54C2KW334147
NA
B 332
6
Q
M
1
0
YTZ-399
F
NA
NA
NA
NA
2G3AJ54N8M2333118
NA
B 333
6
Q
O
1
D
WSZ-617
F
NA
NA
NA
NA
1FABP50D1KG155127
NA
B 334
G
Q
O
1
D
YRZ-186
F
NA
NA
NA
NA
2G4WB54T2M1415297
NA
B 335
3
T
M
1
F
SLC-935
P
N
NA
NA
NA
JN1HB11S5CU006220
NA
B 336
6
Q
O
1
D
YTJ-518
F
NA
NA
NA
NA
1G4AH54N4M6415191
NA
B 337
3
B335
M
1
F
18608HC
P
Y
Y
S
118
JT2MX62EXB0014430
47038
B 33B
3
T
M
1
D
VZH-966
P
Y
Y
S
107
1G6AD478XC9150406
21895
B 339
6
Q
O
1
F
WPG-698
P
N
NA
NA
NA
1YVGD22B9M5105315
NA
B 340
6
Q
O
1
D
WTL-002
P
N
NA
NA
NA
1FABP5346KA149225
NA
B 341
6
Q
O
1
D
BB05062
P
N
NA
NA
NA
1B3XP28D2MN554504
NA
B 342
6
Q
M
1
F
RBM-018
P
N
NA
NA
NA
1YVGD22B6M5125617
NA
B 343
3
B321
M
1
F
ZJH-811
P
N
NA
NA
NA
TE51 -696464
NA
B 344
3
S160
O
1
D
NYJ-276
P
Y
Y
W
126
1Y1SK5167LZ072182
33818
B 345
3
T
M
1
D
XKL-868
F
NA
NA
NA
NA
8Y89A824642
NA
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
VIN Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Confirmed Odomete
by Decoder
B 346
3
B321
M
1
F
VBA-654
F
NA
NA
NA
NA
JT2MXG3E8D0008546
NA
B 347
3
B321
M
1
F
VLY-424
P
N
NA
NA
NA
JM1GD2222J1532651
NA
B 34B
3
B321
M
1
F
VTE-646
F
NA
NA
NA
NA
1N4PB21S1HC885474
NA
B 349
3
B309
M
1
0
1A43459
F
NA
NA
MA
NA
INTERCHANGEABLE
NA
B 350
3
B309
M
1
D
YVJ-165
P
Y
Y
M
142
1P3BB26P3GX509455
69134
B 351
3
S244
M
1
0
PXX-020
P
Y
Y
W
137
1G4AJ47A8EH477206
64321
B 352
3
M
0
217212-M
F
NA
NA
NA
NA
1FMCU14T7JUB25370
NA
B 354
6
Q
M
1
D
WRT-909
Y
Y
NA
BLU7
1FAPP36X0KK176575
42858
B 355
3
M
1
F
WWF-499
F
NA
NA
NA
NA
JT2EL32D3J0254164
NA
B 356
3
M
1
D
F
NA
NA
NA
NA
NA
B 357
6
Q
M
D
391128-M
P
Y
N5
NA
NA
2B4FK25K6KR225698
NA
B 358
6
Q
M
1
D
VXG-781
P
Y
Y
NA
YEL8
1G1AW51W8K6151308
36546
B 359
3
B321
M
1
F
VPS-661
F
NA
NA
NA
NA
1HGCA5620JA0599 66
NA
B 360
3
S238
M
1
D
YZD-267
F
NA
NA
NA
NA
1FABP3191GT119712
NA
B 361
3
T
M
1
F
416-ABJ
P
Y
Y
S
130
JT2A E92W6J307G220
47393
B 362
3
M
1
0
P
N
NA
NA
NA
NA
B 363
6
Q
M
1
D
ZPB0089
P
Y
Y
NA
YEL14
1G1JF14T2M7136097
17769
B 364
3
S096
M
1
D
P
N
NA
NA
NA
NA
B 365
3
B321
M
1
F
PXX-127
Y
Y
S
122
JT2AE83E4G3279918
92930
B 366
3
T
M
1
D
P
N
NA
NA
NA
NA
B 367
3
B366
M
1
D
WVP-17B
P
Y
Y
W
144
1G2JB6903G7538623
72644
B 368
3
S132
Y
1
F
WCC-990
P
Y
Y
S
109
KMHLF31J3JU420717
55521
B 369
3
T
M
1
F
RXY-721
Y
Y
s
143
JT2AE72SXD2040392
56525
B 370
6
a
M
1
F
WXN-807
Y
Y
NA
ORG21
JN1HJ01P7KT248360
31558
B 371
3
B253
M
F
771-871
P
Y
Y
M
117
1N6ND06S0EC332168
93910
B 372
6
Q
M
1
D
XXG-371
P
N
NA
NA
NA
1C3BC6637KD543222
NA
B 373
3
T
0
1
D
PXW-449
P
N
NA
NA
NA
1P3BP26C5DF123152
NA
B 374
6
0
M
1
D
VXL-610
F
NA
NA
NA
NA
1G4AH51N6KT421352
NA
B 375
3
T
M
1
0
TLG-398
P
Y
Y
W
SN08
2G1AW35X3G1114186
94959
B 376
3
B225
M
1
F
SLC-935
Y
Y
S
121
JN1HB11S5CU006220
64020
B 377
3
S096
M
1
D
RTK-925
N
NA
NA
NA
1G3AJ19R5FG398797
NA
B 378
3
M
1
D
VPS-949
F
NA
NA
NA
NA
1FABP55U5KA178079
NA
B 379
3
M
1
F
SRE-517
P
N
NA
NA
NA
JM1GC22A4H1100650
NA
B 380
3
S0S6
M
1
D
NCY-024
F
NA
NA
NA
NA
1B3BD36K4GF152556
NA
B 381
3
S096
M
1
D
PVW-915
Y
Y
139
1G3HY3735G1874913
45727
B 382
3
M
1
F
XBS-139
F
NA
NA
NA
NA
WBADB7401E1194390
No NA
B 383
3
S096
M
1
D
WVP-9B8
N
NA
NA
NA
1FABP5347JA150060
NA
B 384
3
B271
M
1
F
TAH-525
P
N
NA
NA
NA
JT2MX63E2E0054388
NA
B 385
3
M
1
F
VHX-814
F
NA
NA
NA
NA
JN1MN06S1CM004618
NA
B 386
3
B271
M
1
F
XNK-229
P
Y
Y
S
112
JT2AE82E1G3308647
85833
B 387
3
T
M
F
NYN-110
F
NA
NA
NA
NA
JN1PB11SXFU629946
NA
B 388
3
M
1
F
ZJH-796
P
N
NA
NA
NA
JT2RA64LOD0024499
NA
8 389
3
SI 54
M
1
F
YNX-083
P
Y
Y
S
102
JT2TE72W2C5109018
154836
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
VIN
Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Confirmed Odometer
by Decoder
B 390
6
Q
M
1
D
JHU-0078
P
Y
Y
NA
RED1
1P3BA46K1KF430173
36125
B 391
6
Q
O
1
D
PXC-659
P
N
NA
NA
NA
1G2HX54CXKW2G3004
NA
B 392
6
Q
M
1
D
ZJE-033
P
Y
Y
NA
GRN10
2G4WB54T1M180580B
34236
B 393
6
Q
M
1
D
TXC-845
P
N
NA
NA
NA
1G4HR54C1MH474455
NA
B 394
6
Q
M
1
D
005-AHZ
P
N
NA
NA
NA
1G3HN54C2KW324623
NA
B 395
3
M
1
F
092-AKW
P
Y
Y
S
146
JN1PB12S4FU631495
87957
B 396
6
Q
M
1
D
YAC-514
P
N
NA
NA
NA
2G1WL54T3L9124656
NA
B 397
6
Q
M
1
D
SDH-511
P
N
NA
NA
NA
1G1JF11W0K7173824
NA
B 398
3
T
M
1
F
XME-109
P
Y
Y
133
JT2SV21W5H0026106
91220
B 399
6
Q
Y
1
F
WWL-463
F
NA
NA
NA
NA
1N4GB22S1KC768337
NA
B 400
6
Q
O
1
D
BBO-1765
P
N
NA
NA
NA
1G3HN54C0KH300390
NA
B 401
6
Q
O
1
D
XYH-105
F
NA
NA
NA
NA
1G4HP54C1KH536052
NA
B 402
3
T
M
1
D
TPD-264
P
N
NA
NA
NA
1FABR534XJA143314
NA
B 403
6
Q
M
1
D
WNZ-799
F
NA
NA
NA
NA
1G3AJ51W0KG317014
NA
B 404
3
B402
M
1
D
WXF-233
NA
NA
NA
1B3CA44KX JG365292
B 405
6
a
M
1
D
WSA-278
F
NA
NA
NA
NA
1G1AW81WXK6166023
NA
B 406
3
B402
M
1
D
YDT-44B
P
Y
Y
W
120
1G2NE27L1FC748917
92967
B 407
3
T
O
1
D
SDT-657
P
N
NA
NA
NA
1G1JC1110H7162618
NA
B 408
6
a
M
3
D
CWALKER
F
NA
NA
NA
NA
1G1LW14W4KY160895
NA
B 409
3
B300
M
2
D
122140-M
F
NA
NA
NA
NA
D14AB7S218324
NA
B 410
3
S025
Y
2
F
071513
P
Y
Y
S
SN05
JM2UF1135J0369027
69202
B 411
3
T
M
1
F
YTG-562
F
NA
NA
NA
NA
JH4KA765XMC016078
NA
B 412
6
Q
M
1
F
WSW-521
P
N
NA
NA
NA
JP3CU24X2KU050690
NA
B 413
3
B345
M
1
0
NYJ-949
P
Y
Y
S
136
4J47AAG110524
61080
B 414
6
Q
M
1
D
WVD-341
P
Y
Y
NA
YEL9
1G4HR54C3KH422726
55081
B 415
6
Q
O
1
D
WTK-322
P
N
NA
NA
NA
1G4AH51N2KT474159
NA
B 416
3
T
M
1
F
TAJ-048
F
NA
NA
NA
NA
YS3AK46D8L5004586
NA
B 417
6
Q
M
1
D
YNS-238
P
N
NA
NA
NA
1G3CX54C7L4358932
NA
B 418
3
B416
M
1
F
XYK-724
P
N
NA
NA
NA
JN1FU21P4MT313582
NA
B 419
3
S152
M
1
D
SRK-971
P
Y
Y
S
105
4J47WAG128222
68411
B 420
3
T
Y
1
D
TPM-504
P
Y
Y
M
SN04
1G1AB08C3DY262608
89473
B 421
6
Q
M
F
YVC-189
P
N
NA
NA
NA
JM1NA351XM1221420
NA
B 422
6
Q
M
1
D
SBW-055
F
NA
NA
NA
NA
1G3HN54C2KW348470
NA
B 423
3
B416
M
1
F
PMA-698
P
N
NA
NA
NA
JHMEE2754MS003383
NA
B 424
3
M
1
D
XDN-211
P
Y
Y
M
134
1 FABP3191GT147350
57544
B 425
3
S079
M
1
D
P
N
NA
NA
NA
NA
B 426
3
T
M
1
D
ZPJ-808
P
Y
Y
W
131
1B3BS44K1JN276747
77057
B 427
3
M
1
D
VHF-354
P
N
NA
NA
NA
1FABP64W6HH100306
NA
B 423
3
M
1
F
ZHZ-034
P
Y
Y
M
111
JHMAN5523GC000420
103274
B 429
6
a
O
1
D
RTH-131
P
N
NA
NA
NA
2G4WB14T1K1462889
NA
B 430
3
T
M
0
183-090
F
NA
NA
NA
NA
1FTDF15Y4MNA52700
NA
B 431
6
Q
O
1
D
PVZ-471
P
Y
Y
NA
YEL11
1G3HN54C7KH300726
65643
B 432
3
B416
M
1
F
078-AME
P
N
NA
NA
NA
JT2AE94A5M3439124
NA
-------
Table 4-2
(Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
VIN Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Confirmed Odomeler
by Decoder
B 433
6
0
M
1
D
WLP-729
P
Y
Y
NA
YEL13
1G4CW54C6K1617084
17027
B 434
6
Q
M
1
F
YVE-118
P
N
NA
NA
NA
JM1BG2264M0206204
NA
B 435
6
0
M
1
F
894-AAH
P
Y
Y
NA
BLU15
JB3CU14XXKU025841
35500
B 436
3
B416
M
1
D
YWH-584
P
Y
Y
145
4P3CS34TXME109053
No 14984
B 437
3
M
1
0
RMW-619
P
N
NA
NA
NA
1FABP4036FG169788
NA
B 438
3
S079
M
1
F
847-ACG
P
Y
Y
S
104
JT2EL31GOH0069950
45900
B 439
6
Q
M
1
D
TPP-462
P
N
NA
NA
NA
1G3CX54CBK1329939
NA
8 440
3
T
M
1
F
VPS-062
F
NA
NA
NA
NA
JN1HB02S3CV468445
No NA
B 441
3
S146
M
1
D
NKW-779
P
Y
Y
M
101
1G1AB08C5EA125860
45852
B 442
3
T
M
1
D
WYA-251
F
NA
NA
NA
NA
J81RG5192H0458841
NA
B 443
6
Q
M
1
D
YXG-219
P
N
NA
NA
NA
1G4HR54C0KH463458
NA
B 444
3
M
1
F
TPT-481
F
NA
NA
NA
NA
1HGAD7423FA060632
NA
B 445
3
SI 08
M
F
339-997
P
Y
Y
S
108
JT4VR29V5G5007140
157594
B 446
3
M
1
F
WVC-649
F
NA
NA
NA
NA
CM43N6C170081
NA
B 447
3
M
1
D
XBA-668
P
Y
Y
S
140
1G4AM69A7CH145805
92324
B 448
3
B398
M
1
F
WRZ-880
P
N
NA
NA
NA
JT2EL31D9K0349959
NA
B 449
3
T
Ml
1
F
XZC-415
F
NA
NA
NA
NA
KMHLD31J3HU021953
NA
B 450
6
Q
1
D
RTK-269
P
Y
N5
NA
NA
1P3BP44D5KN555699
NA
B 451
3
S233
M
1
D
REV-549
P
Y
Y
W
SN01
1P3BP36K4HF153374
75331
B 452
3
B430
M
D
XXB-029
P
Y
Y
W
123
1GMCU06D2LT200405
No 45317
B 453
6
Q
O
1
D
REM-835
F
NA
NA
NA
NA
1G4H P54C5M H469698
NA
B 454
6
Q
O
1
0
WVH-349
P
Y
Y
NA
YEL12
2G4WD14W3K1400883
45190
B 455
3
M
1
0
XXD-170
F
NA
NA
NA
NA
1G2JB2700F7567426
NA
B 456
3
B440
M
1
F
TXB-267
F
NA
NA
NA
NA
JT2RA63C7F6250542
NA
B 457
6
Q
M
1
D
YPZ-339
P
N
NA
NA
NA
1FACP52U2LA242346
NA
B 458
3
B398
M
1
F
XZH-653
F
NA
NA
NA
NA
4T1SV21E0KU049447
NA
B 459
3
M
1
F
XDF-576
P
N
NA
NA
NA
1HGAD5435E A055864
NA
B 460
3
S246
M
1
0
NJM-372
P
Y
Y
M
SN02
1FABP46F7D4168830
66683
B 461
6
Q
M
1
D
VHK-972
F
NA
NA
NA
NA
1FACP50U9MA194624
NA
B 462
6
Q
M
1
D
64677
F
NA
NA
NA
NA
2FABP74F0KX23O6B9
NA
B 463
3
M
1
D
ZHG-538
P
N
NA
NA
NA
1FABP52U4HG 286736
NA
B 464
3
M
1
0
TPB-641
NA
NA
NA
1G1JC5113JJ176450
NA
B 465
3
M
1
0
TPY-713
F
NA
NA
NA
NA
1G3BN37Y5FY397447
NA
B 466
3
B398
M
1
F
SRY-952
P
Y
Y
S
108
JT2AE72EXD2098772
114401
B 467
3
M
1
F
XTZ-609
P
Y
Y
M
SN03
1HGBA743XGA063308
72248
B 468
3
M
1
F
YVW-272
P
Y
Y
S
113
JT2EL31D5H0075958
141052
B 469
3
B110
M
1
0
YJP-153
P
N
NA
NA
NA
JG1MR6155JKJ02494
No NA
B 470
3
B110
V
1
D
PCB-138
P
N
NA
NA
NA
1P3BP21C4DG136010
NA
B 471
3
B110
M
1
0
VEX-828
F
NA
NA
NA
NA
1G1LV11W5JY205140
NA
B 472
3
B110
M
1
0
ZHB-621
P
Y
Y
S
135
1P3BM18C6ED304919
42090
B 473
3
B117
M
1
D
XSM-185
P
N
NA
NA
NA
1FABP55U8HA18070B
NA
B 474
3
B117
M
1
D
XYK-789
P
N
NA
NA
NA
1FABP22X4GK125291
NA
B 475
3
B117
M
1
D
RTK-159
P
N
NA
NA
NA
1G1AB08C7EY142400
NA
-------
Table 4-2 (Continued)
Sample
Logger
Selection
Driver
Vehicle
Vehicle
License
Vehicle
Driver
Logger
Distance
Logger
VIN
VIN
Final
Number
Candidate
Method
Age
Type
Origin
Number
Screen
Agrees
Installed
Sensor
Number
Confirmed Odomete
by Decoder
B 476
3
B407
O
1
D
RKP-643
F
NA
NA
NA
NA
9K94T231057
NA
B 478
3
B440
M
1
F
AKA-0073
P
N
NA
NA
NA
JT2AE95C1K3212B20
NA
B 479
6
Q
M
1
D
ZFZ-128
F
NA
NA
NA
NA
1B3BC5635KD448505
NA
B 480
6
Q
O
2
D
108029-M
P
Y
N5
NA
NA
2B4FK45K4KR249381
NA
B 481
6
Q
M
2
D
289288 M
P
N
NA
NA
NA
2P4FH45J0KR252752
NA
B 482
3
B407
O
1
D
56936
P
Y
Y
W
SN07
1C3CJ51E3HG185086
23084
B 483
6
Q
M
1
D
VHB-116
P
N
NA
NA
NA
1FACP53UXMG138796
NA
B 484
3
B440
M
1
F
NYL-603
P
N
NA
NA
NA
1HGCA5638JA093906
NA
B 343A
6
Q
M
1
D
YSV-862
P
N
NA
NA
NA
1LNCM83WXMY684364
NA
B 354A
6
Q
M
1
0
YSJ-611
P
N
NA
NA
NA
1G1LT54T7LY262991
NA
N1 VIN not acceptable
N2 No communication with logger
N3 Unable to pick up speed pulses
N5 Error codes
N6 Digital dash
UJ N7 Unacceptable engine family
^ N8 Unacceptable model year
-------
Baltimore City station. Each vehicle was solicited specifically as a candidate for a 3- or
6-parameter datalogger. This is indicated in the Logger Candidate column in Table 4-2.
The method used to select the vehicle from all of the vehicles passing
through the I/M lane is given in Column 3. Selection of 3-parameter dataloggers was
based on a randomly chosen time of day that they passed their I/M inspection. This is
indicated in the table by a "T' in Column 3. Some of the vehicles selected by randomly
selected time were not instrumented. In an attempt to keep the vehicles instrumented in
this study unbiased, vehicles for the 3-parameter datalogger that were not instrumented
were replaced by vehicles with similar vehicle and driver characteristics. These vehicles
are shown in Table 4-2 by a sample number in Column 3 that shows the original sample
number they replaced. For example, sample S011 was successfully instrumented, and it
was a replacement for sample S009, which was solicited earlier but not instrumented.
6-parameter datalogger candidates were not selected randomly and were
not selected according to time but according to a quota system developed to attempt to
meet a distribution of vehicle manufacturers in the study. The selection method for
these vehicles is designated in Column 3 by a "Q".
For the purposes of data analysis, the solicitor estimated the age of the
driver in three categories: Young (<25), Middle (26-65), and Old (>65). The type of
vehicle was placed in three general classes by the solicitor: 1 (sedan, luxury, station
wagon), 2 (pick-up, utility, van), and 3 (sportscar). The origin of vehicle manufacture
was determined and categorized as either F (foreign) or D (domestic). The next column
gives the license number of the vehicle being solicited. In those cases where the license
was an out-of-state license, the state is given in parentheses at the end of the license
number.
Before a vehicle was considered to be an actual participant in the project,
three stages had to be passed: the vehicle had to pass a set of screening questions asked
4-33
-------
of the driver, the driver had to agree to participate, and the datalogger had to be
successfully installed on the vehicle.
During solicitation of the vehicle's driver, several questions were asked to
determine if the vehicle was appropriate for the study. These questions determined if
the vehicle was in acceptable condition for the study, if the driver of the vehicle was the
owner of the vehicle, if the vehicle was driven in normal everyday use, if it was an-
ticipated that the car would be used for any long trips or would have major mechanical
work done during the next week, and if the car was running normally. An undesired
answer to any of these questions caused the vehicle to fail the screening stage. In
neither Spokane nor Baltimore did any vehicle fail the vehicle screen because it was in
unacceptable condition.
In the second stage, the driver was asked if he or she was interested in
participating in the study. At this point, the driver was told that a monetary incentive
was being offered. A "no" answer meant that the vehicle would not be instrumented.
The third stage involved actual installation of the datalogger. In most
cases, the dataloggers could be installed; however, in a few cases mechanical or
electronic problems prevented it from being instrumented. Table 4-2 shows the results
of each of these three stages in the vehicle solicitation and installation process.
The type of distance sensor used for the 3-parameter dataloggers is given
in the Distance Sensor column. S (adaptor on the speedometer cable), W (Scotch-Loked
to the OEM speed sensor wire), and M (magnets attached to a drive shaft) were the
three types of speed sensors used. In Table 4-2 "NA" means that the response is not
applicable. A period indicates a missing value.
Each of the dataloggers was given an identifying code to help ensure that
data from the correct datalogger would be matched with the other information in the
4-34
-------
data packet. This is given in the Logger Number column in Table 4-2. The next column
gives the vehicle identification number (VIN) for most of the vehicles solicited in this
study. The VINs were checked against the 1989 Radian VIN decoder for 1972-1989
vehicles and were checked against VIN decoding literature for 1990-1992 vehicles. For
those vehicles for which a correctly decoded VIN could not be elucidated, Table 4-2
indicates a "no" in the next-to-last column. All other VINs were decoded successfully.
The last column gives the odometer reading of the vehicle when the dataloggers were
removed from vehicles.
The target number of instrumentations for Spokane and Baltimore was 99
3-parameter vehicles and 45 6-parameter vehicles for each city. The 3-parameter
dataloggers were generally left on the vehicles for eight days and the 6-parameter
dataloggers were typically left on the vehicles for seven days to get approximately a full
week's set of data subject to the daily availability of dataloggers ready for installation.
Fifty-five 3-parameter dataloggers and 20 6-parameter dataloggers were available. These
numbers proved adequate to meet the project's needs.
In the following two subsections, the details of the 3-parameter and 6-
parameter instrumentation activities are discussed.
4.3.2 3-Parameter Instrumentation
The 3-parameter dataloggers were designed and fabricated so that they
could be put on any vehicle that passed through the I/M station. This allowed the
selection of vehicles to be instrumented to be as random as possible. The population of
vehicles from which vehicles for 3-parameter dataloggers were solicited were all of those
that passed through the I/M station and that had passed the inspection or been given a
waiver. While in this report these vehicles are called randomly selected, the population
from which they were selected probably does have biases. Only those vehicles that
passed through the inspection station at the time that they were solicited were available
4-35
-------
for instrumentation. Because drivers were offered a monetary incentive to participate, it
is possible that less well-to-do drivers might participate disproportionately in this study.
Table 4-3 summarizes the solicitation and instrumentation success for the
3- and 6-parameter dataloggers in both cities. For example, in Spokane a total of 161
vehicles were solicited for the 3-parameter dataloggers. Of these, 144 passed the
screening questions. Of those that passed the screening, 111 drivers were interested in
participating. Of those interested in participating, the mechanics were able to install
dataloggers successfully on 102 vehicles. Thus, 63 percent of the vehicles solicited for 3-
parameter dataloggers in Spokane were successfully instrumented. In Baltimore the 3-
parameter installation rate was only 36%. Because of this, twice the number of vehicles
had to be solicited in Baltimore as in Spokane.
No vehicle was judged to be in unacceptable condition for this study, and
very few drivers considered their vehicles to be running abnormally. Thus, most of the
failures during the screening phase of the solicitation were due to the driver expecting to
use the vehicle for a long trip, or to have major work done, or because the vehicle was
not their everyday car. Because Washington is a community property state, when the
vehicle was being driven by the husband or wife, the driver was usually the owner of
record for the vehicle. In a few instances, vehicles were being driven by a son or
daughter. Fleet vehicles were eligible for the study; however, usually these vehicles were
driven by an employee, not the owner of the vehicle.
In Spokane, most of the drivers whose vehicles passed the screening
questions were interested in participating in the study; in Baltimore only half of the
screening passes were interested in participating. However, there were some possible
biases in those who refused to participate. Most noticeable were elderly people
4-36
-------
Table 4-3
Summary of Field Activity Rates
City
Datalogger
Type
Vehicles
Solicited
Screen
Passes
Driver
Participations
Successful
Instrumentations
Spokane
3-Parameter
161
144
111
102
Spokane
6-Parameter
85
78
57
42
Baltimore
3-Parameter
318
230
115
113
Baltimore
6-Parameter
163
119
48
37
4-37
-------
who had late-model luxuiy cars. These people tended to not be interested in partici-
pating.
The mechanics were not able to instrument 11 of the 226 eligible cars with
3-parameter dataloggers. Initially in Spokane, the presence of a distributorless ignition
system (DIS) prevented the mechanics from performing an installation because such a
system had not been encountered in the Austin and Spokane pilot studies. However, in
the second half of the Spokane effort, a method for instrumenting these types of vehicles
was found and the vehicles were successfully instrumented.
Table 4-4 identifies the vehicles that had 3-parameter dataloggers installed.
The description of each vehicle was obtained by decoding the VINs. The year, make,
and model of these decoded VINs were checked against the recorded data and the
photograph in the data packets.
The VINs of the vehicles that were not instrumented with 3-parameter
dataloggers were obtained via the vehicle license number. The VINs for these vehicles
were also decoded and are shown in Table 4-5.
4.3.3 6-Parameter Instrumentation
Since the 6-parameter dataloggers were designed to work on certain
models of 1989, 1990, and 1991 vehicles, only those model-year vehicles could be
instrumented. For these instrumentations, the solicitor approached every eligible vehicle,
if possible. Participation in the study was determined by the same three-stage procedure
used to select the vehicles for 3-parameter dataloggers. In addition, 6-parameter
dataloggers could be placed only on certain vehicles because not every late model
vehicle for each manufacturer could be instrumented. Thus, a check of the vehicle
identification number and other vehicle checks (for example, drive train configuration)
had to be made. Finally, the manufacturer-specific datalogger for the vehicle solicited
4-38
-------
Table 4-4
Vehicles Instrumented with 3-Parameter Dataloggers
Sample
VIN
VIN
Mod*)
Manufacturer
Malt*
Modal
Engine
Fuel
Emission
Country
Commanu
Numbar
Confirmed
Year
Induction
Conl^oi
by Daco der
Systamt
S 002
1GCGK24M6GU162254
1986
OA
CHEVROLET
CoflverHonaJ Cab GMT400 4i4
5.7 L
V0
4 bbt
AIR/EVP/TWCEGR^CLL
LISA
TRK 3/4 ton RAT ; 8001 • 9000 Ibt
S 004
KMHLA3IJ8HU121264
1987
HYUNDAI
HYUNDAI
Eacel
15 L
L4
Fl
AIR/EVP/TWOEGR/CLL
KOREA
CAR
5 COS
WAUFC58AXLA061766
1990
AUDI
AUDI
Ouaiiio
21 L
IS TURBO
Ft
Evprrwcca
CAR 20© Turto 4 Dr. Sadan
S 008
1FAPP36X7JK162901
1988
FCFD
FCFD
Twnpo 4 Or. Sedan GL
23 L
L4
CFl
evphwc^gra:ll
USA
CAR
SOU
1G3NT69U9GM372803
1966
OA
OLDSM08JLE
Calais Scprama
25 L
L4
Ff
EVP/TWCVEGRCLL
USA
CAR Sadan 4 Door 4 Window Notdibac*
sots
HL41Q6B367586
1978
CHRYS^R
PLYMOUTH
Volare
318 CI
V8
2 bbl
AIRyfVP/CAT/EGR/
USA
CAR 4 Or. S*dan
SOfl
GCFBAD246990
N>
I960
POflD
PCfD
Fiesta
S 024
JF2KA83A2JD727208
1988
SUBARU
SUBARU
Jutly 4WD
12 L
L3
2 bbl
AIR/EVP/TWCEGR^CLL
JAPAN
MPV Jutty
S025
JT4RN50R6J0353266
1988
TOYOTA
TOYOTA
Land CrUser
23 L
L4
EFI
A JR/E V P/T WCVE G R/C L L
JAPAN
TRK TRK
S 027
39128622
tto
1969
CADILLAC
CADILLAC
Coup* DeVHle
CAR .
S 030
LB110-800914
1972
NISSAN
DATSUN
210
20 L
L4
2 bbl
AIR/EVP/EGRy
JAPAN
CARSadan
S 036
JT3FJ62G5J0067006
1988
TOYOTA
TOYOTA
LandCrulaar
40 L
L«
EFl
AIR/EVP/TWC/EGR/CLL
JAPAN
MPV MPV
S 038
1G4AT27P2EK537437
1984
at
BUtCK
$kyh«wfc LJrolled
20 L
L4
EFI
A JR/E V P/T WC/E G R/C L L
USA
CAR Coup* 2 Door Nofchbac*
S 041
JT2AE92W1J314140S
1988
TOYOTA
TOYOTA
Corolla
15 L
L4
2 bbi
AWEVP/TWC/EGR/CLL
JAPAN
CAR 4*4 Wagon
S 043
1B7FN14C7JS7W743
1988
CHRYS^R
DCCDE
Dodo* Dakota 4X2
22 L
L4
2 bbl
AIR/EVPHWC/EGR/CLL
USA
TRK Conventional Cab RAT : 04001-05000 GVWft 6
S 046
2FTEF14N6QCA49786
1986
POPO
FCFD
F1M Uflhl True*. 4K4
50 L
V8
EFI
AIR/EVP/TWC'EGR/CLL
CAHADA
TRK Ught Truck Rag Cab RAT : 6001 • 7000
S 048
AA0BE8S328524
197B
CHRYSLER
PLYMOUTH
PW100 - Plymouth Trail Dustar
318 CI
V8
2 bbl
EVPyCAT/EGR/
USA
TRK Sport Utility RAT : 6000# to 10000#
S 050
1MEBP89C9EG646134
1984
FCFD
MEROJW
Mar quia 4 Dr. Sedan LTS. Brougham
36 L
V6
2 bbl
A1R/EVP/TWC/EGR/CLL
USA
CAR
S 054
1B7HW14T0GS016S4S
1986
CHRYSLER
DCOGE
Dodge Ram Pkkt*> 4X4
52 L
V8
2 bhd
AIR/EVP/CAT/EGR/
USA
TRK Conventional Cab RAT : 06001-07000 GVWR &
S 055
1J089AA196186
1990
94
CHEVROLET
Chevelte Scooter
16 L
L4
2 bU
AIR/EVP/CAT/EGR/
USA
CAR Sedan 2 Door Hatchback
S 061
JM 1BF222700145368
1986
MAZDA
MAZDA
323
16 L
L4
Fl
EVP/TWC/EGftCLL
JAPAN
CAR 4 O. Sedan
S066
1FABPJ934EG169705
1984
PDFD
FQFO
LTD Brougham HT 4 Dr Sedan
36 L
V6
2 bbl
A«R/EVPrrwC/EGR/CLL
USA
CAR
S 067
1C36H58EIGNI68729
1986
OflYSLER
CNRYS^R
La Baron, QTS
2 2 L
L4 TURBO
2 bt*
EVP/TWC/EGR/CLL
USA
CAR 4 Or. Hatchback
S 069
JM 1BF2327J011697S
1986
MAZDA
MAZDA
323
16 L
L4
Fl
EVP/TWC/CU
JAPAN
CAR 3 Or. Hatchback
S 072
2G2AK37HSE2268581
1984
at
PONT'AC
Grand Prix L£
SO L
V6
4 bbl
AWEVP/7WC/EGR/CLL
CANADA
CAR Coup* 2 Door Notdibadt Sped*)
S 073
JF2AN55B4GD439011
1986
SUBARU
SUBARU
Turbo
18 L
H4
2 bbl
WR^EVP/TWC/EGRyCLL
JAPAN
MPV Station Wagon
S 076
1FABP22 X2GK155745
1986
R3FD
RH)
Laser 4 Dr Sedan
23 L
L4
CPI
A1R/EVP/TWOEGR/CLL
USA
CAR
S 076
1C3BAS4E0EG237B12
1984
&RYSLER
ORYSLER
Laser
22 L
L4 TURBO
2 bbl
EVP/TWC/EGR.'CLL
USA
CAR 2 Or Hatchback
S 079
1B30Z18C5GO231O63
1986
OflYSLER
DCOGE
OMN
22 L
L4
2 bbl
EVP/TWC/EGR/CLL
USA
CAR 4 Dr. Hatctoaefc
S 086
JT2AE9ZE6J3034661
1986
TOYOTA
TOYOTA
Corolla
15 L
L4
2 bbl
AlfVEVP/TWC/EGfl/CLL
JAPAN
CAR 4 Dr Sedan
S 090
1GNDL15Z6L0143279
1990
OA
CHEVROLET
Astro 4*4
4 3 L
V6
Fl
AIR/EVP/TWC/EGR/CLL
U.SA
MPV 1/2 ton RAT : 5001 • 6000 lbs
S 094
1G2WKI4W3JF252210
1986
OA
PONDAC
Cfrand Prix LE
28 L
V6
Fl
A WE V P/TWC/EG R/C L L
USA
CAR 2 Door Coupe/Sadan
S 095
1MEBP75X7GK641646
1966
POflD
MBCUW
Topaz 4 Or Sadan GS 4 WDR
2 3 L
L4
CFI
AWEVP/TWC/EQR/CLL
USA
CAR
S 109
1G1AW19R0G6112673
1986
OA
CHEVROLET
Calabrlty
25 L
L4
Fl
EVPiTWC/EGR.'CLL
USA
CAR Sedan 4 Door 6 window Nottfibaefc
S 103
1G1LV14W9JE673479
1968
OA
CHEVROLET
Beralla
2 8 L
V6
Fl
AWEVP/TVKVEGR/CLL
USA
CAR 2 Door Coupe/Sedan
S 105
1FABP0521CW107600
1902
pony
FOFD
Escort 3 Or. Sadat Hatchback
16 L
L4
2 bbl
AIR/EVP/CAT/EGR/
USA
CAR
S 108
JN6ND16Y0GW003899
1906
NISSAN
NISSAN
Truck
24 L
L4
EFI
AlfUEVP/TWC/EGR/CLL
JAPAN
TRK Pathfinder
S 110
JB3BA26KXGU135565
1986
CJflYS^R
DODGE/MTS
Coll
15 L
L4
2 bbl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR SUfon Wagon {Import)
S 112
1YVGD31BXL5201264
1990
MAZDA
MAZDA
MX6
22 L
L4
Fl
EVP/TWC/EGR/CLL
USA
CAR 2 Dr Couoe
S 114
KMHLF21J9HU104162
1967
HYUNDAI
HYUNDAI
Exeat
15 L
L4
Fl
AIR^VF/TWOEGR/CLL
KOREA
CAR
$ 117
JT2AE94A4L3346545
1990
TOYOTA
TOYOTA
Corolla
15 L
L4
2 bbl
AIR/EVP/TWC/EGR^CLL
JAPAN
CAR
S MS
1G8CT18B1E0141922
1984
04
CHEVROLET
SmaH Conventional Cab 4x4
28 L
V6
2 bW
AIR/EVP/CAT/EGR/
USA
MPV 1/2 ton RAT : 4001 • 5000 lbs
S 120
J T2SV22E0L 3407469
1990
TOYOTA
TOYOTA
Camry
2 0 L
L4
EFI
EVP/TWC/EGR/CLL
JAPAN
CAR 4 Dr. Sedan
S 122
JHMSM5423BC116394
1981
HONDA
HONDA
Accord
1.7 L
L4
2 bW
EVP/CAT/EGA/
JAPAN
CAR 4dr Sedan
S 126
1Y1 SK5467LZ153033
1990
a*
CHEVROLET
Geo Prion
16 L
L4
Fl
EVP/TWC/EGR/CLL
USA
CAR 4 Door Sedan
S >26
1NXAE82G3JZ543668
1980
TOYOTA
TOYOTA
Corolla
15 L
L4
2 bbl
AIR/EVP/TWCEGR/CIL
USA
CAR 3 Dr Liflbac*
S 130
1FABP0757EW169481
1964
FORD
KPD
Escort 2 Dr. Sedan Hatchback GT
1.6 L
L4
2 bW
AIR/EVP/TWCEGR/CLL
USA
CAR
S 132
JHMAF5332ES007287
1984
HONDfc
HONDA
OvkCRX
15 L
L4
2 bbl
Ain/EVP/CAT/TWC/EGR/CLL
JAPAN
CAP 2 Dr Hatchback/Coup a
S 134
1HGCA554XJA1056C0
1968
HONCA
HONDA
Accord
2.0 L
L4
2 bbl
EVP/TWC/CLL
USA
CAR 4dr Sedan (man)
S 135
JB7FM55E2JP049655
1986
CHRYSLER
DODGE/MI TS
Dodge Ram pk*-up 4X4
2 5 L
L4
TBI
AIR/EVP/TWCEGR/CLL
JAPAN
TRK Club Cab RAT : 04001-05000 GVWR & H
S 137
1G2PG9793GP263091
1986
OA
PONT1AC
Hero GT Coupe
2 8 L
V6
Fl
A1R/EVP/TWC/EGR/CLL
U.SA
CAR Coupe 2 Door Nolchback Sport
S 136
3M69FAR4 76562
1900
OA
0LDSM06ILE
CtAass Supreme Broujpiam
4 3 L
V8
2 bbi
AIR/EVP/CAT/EGR/
USA
CAR Sedan 4 Door Nolchback
S 143
JM 1BF2321J0174161
1968
MAZDA
MAZDA
323
1.6 L
L4
Fl
EVP/TWC/CLL
JAPAN
CAR 3 Dr. Hatchback
S 144
1FABP28A4DF 126103
1963
FOFD
POPO
Mustang ti 3Dr. Hatcttoecfc
2 3 L
L4
2 btt
AIR/EVP/TWC/EGRfCLL
USA
CAR
$ 146
1FABP4634EH101344
1964
FCFD
POPO
Thindertwd 2 Dr. Sedan
3 6 L
V6
2 bW
AIR/EVP/TWC/EGRCLL
U.SA
CAR
S 147
JM2UF11 33J0354042
1960
MAZDA
MAZDA
B2200 True*
2 2 L
L4
2 bbl
EVP/TWC/EGR/CLL
JAPAN
TRK Short Bed
S 152
1G3CX69B6G1323859
1966
OA
OLD5MOBILE
Ninaty-Eigtii Regency
3 6 L
V6
Fl
AIR/EVP/TWC/EGR«IL
USA
CAR Sedan 4 Ooor 4 Window Notchbadi
S 154
SMG-2023538
1900
HONDA
HONDA
Accord
18 L
L4
3 bbl
EVP/
JAPAN
CAR 3 Door LX
$ 156
1ZVP720C1L5138646
1990
POPO
FORiyMAZDA
Piob* GL 30r Sedan
2 2 L
L4
EFI
EVP/TWC/EQft/CLL
USA
CAR
S ISA
1B3BA64E 3EG120750
1904
CHRYSLER
DCCGg
Oeytona (U.S.}
22 L
L4 TURBO
2 bbl
EVP/TWC/EGR/CLL
USA
CAR 2 Dr. Hatchback
S 160
1FABP52UXJG185500
190B
FCFD
FCFD
Taurus GL 4 Dr. Wagon
3 0 L
V6
EFI
EVP/TWC/CLL
USA
CAR
-------
Table 4-4 (Continued)
S«mpl« VIN
Numb a r
try Decode'
4*
¦U
o
S 163
S 165
S 167
S 169
S 172
S 175
S 176
S 178
S 179
s tec
S 184
S 187
s ies
S 194
S 196
S 197
3 200
S 203
S 204
S 20S
S 207
S 209
S 211
S 212
S 214
S 216
S 219
S 221
S 222
9 224
S 226
S 227
S 229
S 233
S 235
S 238
S 239
S 240
S 241
S 244
5 246
8 002
B 007
B010
B 016
60(7
B 019
B 037
B 039
B 042
B 046
B 050
B 052
B 056
B 059
B0«2
B 0*6
B 072
6 079
B 061
B 004
1GCEK14L5EJ110666
1N4GB22B9LC742425
1QCBS14E3Q21404S6
MX9424JQ3166S9
2B4FK41K2JR509231
JN1PS26S4EW63J444
256690C110112
JF1AC43BUC23049J
1GCBSI4ESQ2149162
JM2UC2217C0SS4694
RA42-061277
1G8CT16R5G6134626
1FAPP23J6JW186397
JM2UF21! t(30523105
U1SSLFE3261
RN47-029466
1X11SA6114134
1FABP34970W162154
5372057163
JM2UF 111900634121
JN1HT14S3DT1O3804
17A0967B15
JM1BF 2226G01534 32
1M07VA7159960
1N6HD11YXJC317900
SMK-2089776
1W27M6K407920
1G1A035PS6 J104025
JM2UF1136K0765114
1FABP64 T6JH140673
1FTBR10C9JUB99316
1FABP40A4JF228320
JM2UC1216E0624500
IP 3BS48KXJN f4 3063
2G2AF51R0H9223691
1C3CJ41E6JG327477
1G1JF77W1GJ208794
JT2ST65L5G7024536
1FMDU15Y1ELA53606
1FA8P55 1/6JG f 22178
1FABP52D3JG173296
1FTCR10A1MUC46371
JHMfiB 7232GC026142
YV1AX665XJ1793960
1G1FP2 J S3KL104669
YS3AL75L6M7906092
1G1FP21SUL193166
2P4FH413XJR662190
1P3BK46D2KC469869
1G6AD6963D9266654
1N6S011S6MC323930
1GCEG25HXC7125742
1HQCB 7661M A04 9561
1G2AB2705E 7332521
1G1JF11W2K7160967
IFAPP14J2MW269094
1YVGD31B5M5138706
1FTCR10A1KUA84004
1FTBR10T6GUB61390
1G1JC1113KJ171405
1G2AB2707E7250564
Mode)
Manufacturer
Make
Modal
Engine
FO0l
Emlaaion
Country
Year
Induction
Control
Systems
1964
CM
CHEVROLET
Convenflonal Cab GMT400 4*4
5.7 I
V0
4 btt
AIR/EVP/TWC/EGR/CLL
USA
1990
NISSAN
NISSAN
Sentra
16 I
L4
2 bU
USA
1986
OA
CHEVROLET
SmaH Conventional Cab
25 L
L4
Fl
EVP/TWC/EGR/CLL
USA
1980
SUBARU
SUBARU
XT-6
2 7 L
H6
Pi
EVP/TWC/CLL
JAPAN
1960
CWYSLER
fYTry
Dodge Caravan
25 L
L4
EF1
AIR/EVP/TWC/EGR/CLL
CANADA
1984
NISSAN
NISSAN
200SX
16 L
L4 TURBO
Fi
EVprrwC/BQfVCLL
JAPAN
1970
<3A
PONT1AC
Executive 400
1986
SUBARU
SUBARU
a.
18 L
H4
2 bbt
EVPflWC^OfVCLL
JAPAN
1986
at
CHEVROLET
Small Conventional Cab
25 L
L4
Ft
evp/twc^gra:ll
USA
1962
MAZDA
MAZOA
B2000 Truck
22 L
L4
2 bbi
AIR/EVP/CAT/EGR/
JAPAN
1976
TOYOTA
TOYOTA
CeNca
20 L
14
2 bbf
AIR/EVP/EGR/
JAPAN
1966
OA
CHEVROLET
Small Conventional Cab 4*4
26 L
V6
ft
AtfvevprrwaEOR/cLL
USA
1966
RDFD
FCFD
Escort 2 Or Sedan Hatchback GT
19 L
L4
EF1
air/evp/twc;egr/cll
USA
1966
MAZDA
MAZOA
B200Q Truck
22 L
L4
2 t>U
AIR/EVP/TWC/EGR«LL
JAPAN
1976
R*£>
FCFD
fronoo 4X4
400 CI
V6
2 bbl
EVP/
USA
1974
TOYOTA
TOYOTA
Hi-Lu*
20 L
L4
2 bW
EVPJEGPV
JAPAN
1960
OA
CHEVROLET
Citation
25 L
L4
2 bbl
EVPiTYiC^OFVCLL
USA
1966
POFC
FORD
Elcort 4 Or. Sfalon Wagon L
19 L
L4
2 bU
AIR/EVP/TWC/EGR/
USA
1977
VW
VW
Sdroeeo
16 L
L4
FI
EVP/EGR/
1966
MAZDA
MAZDA
B2000 Truck
22 I
L4
2 bbi
air/evp/twc;egr/cll
JAPAN
1963
NISSAN
OATSUN
Stent*
19 L
L4
Ft
AIR/EVP/TWCJEGR/CLL
JAPAN
1960
VW
VW
Rabbit
16 L
L4
Ft
EVP/CAT/6GR/
OBWW
1966
MAZOA
MAZOA
323
16 L
L4
PI
EVP/TWC/EOR/CLL
JAPAN
1960
OA
CHEVROLET
Morua
25 L
L4
2 btt
EVP/TWC/EGR^LL
USA
1966
NISSAN
NISSAN
Truch
30 L
V6
EFI
AIR/EVP/TWDEGR/CLL
USA
1960
HONDA
HONDA
Accord
16 L
L4
3 btt
EVP /
JAPAN
1976
OA
CHEVROLET
Maiibu
33 L
V6
2 bU
EVP/CAT/EGR/
USA
1964
Of
CHEVROLET
Cevaffer C9
20 L
L4
EFI
AIR/EVP/TWC/EGR/CLL
USA
1969
MAZOA
MAZDA
B2200 Truck
22 I
L4
2 b W
EVP/TWC/EGR^LL
JAPAN
1966
FCflD
FCRD
Thunder bird Tubo Cpe 2 Dr. Sedan
23 L
14 TURBO
en
AIRIEVP/TWCJEGR/CLL
USA
1966
PCPD
FCFC
Ranger Super Cab
2.0 L
L4
FI
EVP/TWC/EGR/CLL
USA
1966
POFD
FCPD
Mustang LX 2 £>. Sedan
2 3 L
L4
1 &W
EVP/TWC^GRCLL
USA
1984
MAZOA
MAZDA
B2000 Truck
22 L
L4
2 bbt
EVP/TWC/EGR/CLL
JAPAN
1986
CHRY9LER
PLYMOUTH
Sundance
25 1
L4
1 bb)
EVP/TWC/EGR/CLL
USA
1987
OA
PONT) AC
Pontic 6000
25 i
L4
EVP/TWC/EGR/CLL
CANADA
1986
cmvstfR
ORYSLER
Le Baron
22 L
L4 TURBO
2 Bbi
EVP/TWC/EGR/CLL
USA
1986
OA
CHEVROLET
Cavalier 224
28 L
V6
FI
air/evp/twc;egr/cll
USA
1986
TOYOTA
TOYOTA
Cellca
20 L
L4
En
EVP/TWC/EGR/CLL
JAPAN
1984
FCRD
FCPD
&onco 4X4
4 9 L
16
4 btt
AtR/EVP/CAT/EGR/
USA
1986
FCflD
ROT
Taurua L 4 Or Wagon
30 L
V6
EFI
EVP/TWC/CLL
USA
1986
PCPD
PCFD
Taunja GL 4 Dr. Wagon
2 5 I
L4
cn
AIR/EVP/TWC/EGR«LL
USA
1991
FCPD
FCPD
Ranger Super Cab
23 1
14
EFJ
EVP/TWC/EGR/CLL
USA
1986
H0M*
HONDA
Prelude SI
20 L
L4
EFI
EVP/TWC/EGR/CLL
JAPAN
1986
VOLVO
volvo
240 Sedan 6 Wagon
23 L
L4
Ft
EVP/TWC/CLL
9WECCN
1989
CM
CHEVROLET
Camaro Sport
2 8 L
V6
FI
EVP/TWC/EGR/CLL
USA
1991
SAAB
SAAB
900 Turbo SeHea
2.0 L
L4 TURBO
FI
EVP/TWC/CLL
3WECEN
1986
OA
CHEVROLET
Camaro Sport
2 8 L
V6
FI
AIR/EVP/TWC/EGR/CLL
USA
1986
Ctf=fYSL£R
PLYMOUTH
Ptymouft Voyager
30 L
V6
EFI
AIR/EVP/TWC/EGR/CLL
CANADA
1989
CHRYSLER
PLYMOUTH
Reliant, Cuatom, SE, LE
22 L
L4
FI
EVP/TWC/EGR/CLL
USA
1983
OA
CADILLAC
De Vide
4 1 L
V6
DF1
AIR/EVP/TWC/EGR/CLL
USA
1991
NISSAN
NISSAN
True*
24 L
L4
2 bbi
AIR/EVP/TWC/EGR/CLL
USA
1982
OA
CHEVROLET
Van
5.0 L
V6
4 bbf
EVP/CAT/EGR/
USA
1991
HONDA
HOfCA
Accord EX 4 Or Sedan
2.2 L
L4
FI
EVP/TWC/EGR/CLL
USA
1984
OA
PCVTIAC
Synblrd 2000
16 L
L4
TBI
AIR/EVP/TWC/EGR/CLL
USA
1989
OA
CHEVROLET
Cavalier Z24
26 L
V6
FI
EVP/TWC/EGR/CLL
USA
1991
PCPD
FCPD
E»cwt LX 4 Or. Hatchback
1 9 L
L4
En
EVP/TWC/EGR/CLL
USA
1991
MAZDA
MAZOA
MW5
22 I
L4
FI
EVPfTWC/EGR/CLL
USA
1989
PCPD
FCPD
Ranger Super Cab
23 L
14
en
EVPrrwC/EGfVCLL
USA
1986
pert)
FCRD
Ranger Super Cab
29 1
V6
EFI
EVP/TWC/EGR/CLL
USA
1989
OA
CHEVROLET
Cavalier
20 I
L4
FI
EVP/TWC/EGR/CLL
USA
1984
OA
PONTtAC
Sunbird 2000
1.0 L
L4
TBI
AtR/EVP/TWC/EGR/CLL
USA
TRK 1/2 ton RAT : 6001 • 7000 Ibt
CAR 2(k Sedan
TRK 1/2 ton RAT : 3001 - 4000 lb«
CAR XT Coup*
MPV Wagon RAT : 04001 05000 GVWR & H
CAR 2*2 (300ZX)
CAR
CAR 4 Or. Sedan
TRK 1/2 ton RAT : 3001 - 4000 Ibe
TRK4 Or. Sedan
CAR 2 Or Coup«
MPV 1/2 ton RAT : 4001 - 5000 IDS
CAR
TRK Long Bed
TRKfronco RAT: <6000
CAR 2 Dr. Pick-up
CAR Coup* 2 Door Sport
CAR
CAR
TRK Short Bed
CAR 2tk Hardtop
CAfl
CAR 4 Or S#d«n
CAR Coupe 2 Door Special Hatchbat*
TRK Regular Bed
CAR 4 Door
CAR Coup® 2 Door Noldrbac*
CAR Station Wagon 4 Door
TRK Short Bed
CAR
TRK Ugm Truck Rangar RAT : 3001 • 4000
CAR
TRK Short Bed
CAR 4 Or. Hatchback
CAR 4 Door Sedan
CAR 2 Or. SedafVCoype
CAR Coupe 2 Door Plain Bach Hatchback
CAR 3 Dr. Uftback
VAN Bronco RAT : 5001 • 6000
CAR
CAR
TRK Ught True* Rangar RAT : 4001 • 5000
CAR 2 Door
CAR 5 doort
CAR 2 Door Halchback/Ultback
CAR ConvertrWe
CAR 2 Door Halchbac*
MPV wagon RAT : 04001-05000 GVWR 8, H
CAR 4 Or. Sedan
CAR Sedan 4 Door Notchbacfc
TRK Regular Bed
TRK 3/4 ton RAT : 6001 • 7000 lbs
CAR 4dr Sedan (auto)
CAR Coupe 2 Oo-of Nolchback
CAR 2 Door Coupe Sedan
CAR
CAR 2 Or Coupe
TRK U9K True* Ranger RAT : 4001 • 5000
TRK Ught Truck Ranger RAT . 3001 - 4000
CAR 2 Door Coupe Sedan
CAR Coupe 2 Door Notchback
-------
Table 4-4 (Continued)
Sample
VIN
VIN
Model
Manufacturer
Make
Model
Engine
Fuel
Emission
Country
Comrnerls
Number
Contirmefl
Year
Induction
Control
by Decoder
Syjlems
B 086
IFAPP36X9KK201666
1969
FCFD
FCFD
Tempo 4 Dr Sedan OL
2 3 L
L4
CF1
EVP^TWC/EQR/CLL
USA
CAP,
B066
1GCOC14Z6KZ211565
OA
CHEVROLET
Conventional Cab GMT400
4.3 L
V6
Fl
AIFVEVP/TWC/EOR/CLL
USA.
TRK 1/2 ton RAT S001 - 6000 rbs
8 096
1GTBSI4R2G2533452
1986
oa
owe
Small Conventional Cab
2 6 L
V6
Fl
AIWEVP/TWC/EGR/CLL
USA.
TRK 1/2 ton RAT : 3001 • 4000 bs
0 104
(Q3AJII37HD381642
34
OLDSM06H£
Cutlass Oera LS
36 L
V6
Fl
AlfVEVPrrwC/EQR'CLL
USA.
CAR 2 Door Coupe/Sedan
B 105
JT4RN50R7G0173173
1966
TOYOTA
TOYOTA
Pic* up
23 L
L4
EFI
A(fVEVP/TWOEQR/CLL
JAPAN
TRK TRK
B no
1H24KA4409308
1980
OflYSLER
PLYMOUTH
Champ
16 L
L4
2 bW
Alfl/EVP/CAT/EOR/CLL
USA.
CAR 2 Dr Hatchback
B 117
101JC6116H7116530
OA
CHEVROLET
Cavalier
2 0 L
L4
Fl
EVP/TWC/E3R/CLL
USA.
CAR 4 Door Station Wagon
B (23
TE72-509680
Mi
1960
TOYOTA
TOYOTA
Corolla
1.6 L
L4
2 bU
AfR/EVP/CAT/EQR/
JAPAN
CAR 2 Dr/4 Dr Wagon
B 124
1B3BP4605KN589752
OffTfSLER
OCCGE
Shadow
22 L
L4
Fl
EVP/TWC/EGRiCLL
USA.
CAR 4 Dr. Hatcftbacfc
B 130
1FABP44E2KF117996
FCFD
FCFD
Mustang LX 2 Or Sedan Convarfble
50 L
V6
EFI
AIR/EVP/TWC/EGR/CLL
USA.
CAR
B 139
1G4ALI937F6463396
1965
<34
BUICK
Century Unified
3 6 L
V6
MR
AIR'EVP/TWC/EGR/CLL
USA
CAR Sedan 4 Ooor 6 Window NotchbacK
0 U3
JT2AA44C6BOOOI018
1961
TOYOTA
TOYOTA
Cellca
2 4 L
L4
2 bH
AWEVP/TWC/EOR/CLL
JAPAN
CAR 2 Or Coupa
B ISO
1B3YA44K6JG446546
CHRYSLER
DCOGE
Daytona (U.S.)
2 5 L
L4
1 bU
EVP/TWC/EGRCLL
USA.
CAR 2 Dr. Hatchback
B 153
1GNCT18Z3K0152536
1989
OA
CHEVROLET
Small Conventional Cab 4*4
43 L
V6
Fl
AIR/EVP/TWC/EGR/CLL
USA.
MPV 1/2 ton RAT : 4001 - 6000 lbs
B 1S6
J3CFC24H1KE115676
CM
CHEVROLET
Conventtonal Ctb QMT400
50 L
V6
Fl
EVP/TWC/EGR^CLL
USA
TRK 3/4 ton RAT : 7001 • 8000 lb«
B 163
JHMCA5386JC052224
HONDA
HONDA
Accord
2.0 L
L4
EFI
EVP/TWC/CLL
JAPAN
CAR 2dr Hatchback (man)
3 164
103BD31O6FG241191
1985
CHRYSLER
DODGE
AriM SE
2 2 L
L4
Fl
EVP/TWC/EGR/CLL
USA
CAR 2 Dr. SedarVCoupe
B 160
1G4XB59R0FW454358
1985
CM
BUJCK
Skylark Custom
25 L
14
EFI
EVP/TWC/EGR/CLL
USA
CAR Sedan 4 Door 4 Window Notchbac*
0 170
JH4DA34S6KS013350
HONDA
ACORA
Integra 3dr
16 L
L4
EFI
EVP/TWC/CLL
JAPAN
CAR Integra 3dr HatchbacK 4ip«d
B 164
1G4AH51RIHT438622
1987
CM
BUfCK
Century Custom
2 5 L
L4
Fl
EVP/TWC/EGR'CLL
USA.
CAR 4 Door Sedan
B 190
JT2EL34B2M0027569
TOYOTA
TOYOTA
Tercel
1.5 L
L4
2 bW
AlWEVP/TWC/EGR/CLL
JAPAN
CAR
B 191
1GCDC14H3ME100424
1991
CM
CHEVROLET
Conventonal Cab GMT400
5.0 L
V6
Fl
EVP/TWC/EGR/CLL
USA
TRK 1/2 ton RAT : SO01 • 6000 lbs
B 193
1GCOCI4Z6KE135013
1869
OA
CHEVROLET
Conventional Cab QMT400
4 3 L
V6
Fl
AtfVEVP/TWOEGR/CLL
USA.
TRK 1/2 ton RAT : 5001 - 6000 lbs
B 195
JHMBA743 IOC066566
HOfC*
HONDA
Accord
20 L
L4
EFI
EVP/TWC/EGR/CLL
JAPAN
CAR 4 Door Sedan
B 206
2MEBM7SF1KX665S83
FCFD
VBKtfff
Grand Marqiis 4 Or Sedan LS
5.0 L
V8
2 bW
AfWEVP/TWO-EGR/CLL
CANADA
CAR
B 207
JGf MR2152HK74645S
OA
SUZUKI
Sprint
J.O L
L3
2 bU
AIWEVP/TWC/EGR/CLL
JAPAN
CAR 2 Ooor Hatchback
B 206
JM1FC33I2G0I21078
1086
MAZDA
MAZDA
RX7
13 L
Rotary
H4
Fl
AtR/EVP/T WC/CLL
JAPAN
CAR Coupe
8 210
JFIAC42B2KC210379
SUBARU
SUBARU
CL
16 L
2 bW
EVP/TWC/EGfVCLL
JAPAN
CAR 4 Dr Sedan
B 217
tFABPI 3230T120604
1683
POfO
FCFO
Eccort 4 Or Sedan Hatchback
16 L
L4
2 bU
A1R/EVP/TWC/EGR/CLL
USA
CAR
B 233
JN1GB21SXKU537381
NISSAN
NISSAN
Santr*
1.6 L
L4
Fl
AIR/EVP/TWC/EQR«LL
JAPAN
CAR 4dr Sedan
8 236
IW19J9B567918
CM
CHEVROLET
Mallbu
4.4 L
ve
2 bbl
EVP/CAT/EGR/
USA
CAR Sedan 4 Door Notcttucfc
B 241
1MEBM55UXK A633641
TCFD
MTOLWf
Sable 4Or. Sti W71 OS
30 L
V6
EFI
EVP/TWC/EGRiCLL
USA
CAR
B 243
YV1AX8852 HI 723647
VOLVO
VOLVO
240 Sedan 6 Wagon
2.3 L
L4
Fl
evp/twocu
SWEDEN
CAR 5 doora
B 247
2B4FK5 131JR676116
CHRYSLER
DCOGE
Podge Caravan
3.0 L
V6
EFI
AIR/EVP^TWC/EGR/CLL
CANADA
MPV Wagon RAT : 04001-05000 GVWR 8 H
B 255
1C3BC59KXHF330954
OfTYSLEfl
CHRYSLER
Town and County
26 L
L4
1 btt
EVP/TWC/EGR/CLL
USA
CAR SUIon Wagon (Import)
6 256
2FABP38S0HB102197
R0PD
FuF®
Tempo Sport 4Dr. Sedan QLS
23 L
L4
CR
EVP/TWC/EGR/CLL
CANADA
CAR
8 263
JHMBA3144H C019202
1)67
HOCA
HONDA
Prelude SI
20 L
L4
EFI
EVP/TWC/EGR/CLL
JAPAN
CAR 2n»ty.Eight Regency
36 L
V6
Fl
AIR/EVP/TWC/EGR/CLL
USA
CAR Sedan 4 Ooor 4 Window Notcftbadi
B 315
1FTEX15H2GKA94158
1966
POFO
fcfd
F150
5 6 L
V6
4 bW
EVP/
USA
TRK Light Truck Super Ca RAT : 6001 • 7000
B 317
6G87H169118
PCFO
R3FD
Thunderblrd HT SOr.
351 CI
V6
2 bW
AIR/EVP/CAT/EGR/
USA
CAR
B 319
1FABP259XHW116408
raflD
POFD
Escort 4 Or. Sedan Hatchback GL
1 9 L
14
2 bW
EVP/TWC/EGR/CLL
USA
CAR
B 325
JT4RN50R7HO325017
TOYOTA
TOYOTA
Land Cruiser
2 3 L
L4
EFI
AIR/EVP/TWC/EGR/CLL
JAPAN
TRK TRK
B 329
tQ1Q211Q2HP 118178
1967
CM
CHEVROLET
Monte Carlo S
5 0 L
ve
4 bbi
AWEVP/TWC/EGR/CLL
USA
CAR 2 Door Coupe/Sedan
B 337
JT2MX62EX0OO1443O
TOYOTA
TOYOTA
Crestlda
26 L
L6
EFI
EVP/TW
-------
Table 4-4 (Continued)
Sampl*
VIN
VIN
Mod*
Manufacturer
Mafca
Modal
Englna
Fuat
Numtotf
Confirmed
Y «ar
Indue
try Dacocfer
B 367
1G2JB6903G 7536623
1986
CM
PONTIAC
SonbJfd
18 I
L4
Ft
B 366
KMHLF31JiiU42Q7t7
1986
VfYUNDA)
MYUNCW
E*caf
15 I
14
F1
B 369
JT2AE72SX0204Q392
1983
TOYOTA
TOYOTA
Corolli
15 L
L4
2 bbl
8 37 1
1 N6ND06S0EC332168
1964
NISSAN
NISSAN
Truck
24 L
14
2 bbt
B 375
2G1AW35X3Cm 14186
1986
(W
CHEVROLET
Calabrlty
26 I
V6
2 bbt
B 376
JN1HBMS5CUO0622O
1962
NISSAN
DATSUN
Santra
15 I
L4
2 bbt
8 381
t03HY3735Q 1874913
1906
CM
OLDSM0B1LE
Datla 66 Royala Brougham
36 L
V6
Fl
B 386
JT2AE02E 103308647
1966
TOYOTA
TOYOTA
Corolla
15 L
L4
2 tbi
B 389
JT2TE72W2C5IO9018
1982
TOYOTA
TOYOTA
Corolla
16 L
L4
2 Ctf
B 30$
JN1PB12S4FU631495
1965
NISSAN
NISSAN
Santra
16 L
L4
2 bbl
B 398
JT2SV21W5H0026108
1987
TOYOTA
TOYOTA
Camry
20 L
L4
ER
B 406
1G2NE27L1FC746917
198S
OA
PONTIAC
Grand Am
30 L
V6
MB
B 4 10
JM2UF1 1 3SJ0369027
1966
MAZDA
MAZDA
02200 True*
22 L
L4
2 btrf
64(3
4J47 AAG110524
1960
CM
BUCK
36 L
V6
2 bW
B 419
4J47W A0126222
1080
OA
BUCK
Ragal
49 L
V6
4 bbl
B 420
1Q1ABO8C3DY2626O0
1983
OA
CHEVROLET
Chavatta
16 L
L4
2 bbt
B 424
1FABP3191GTM7350
(986
KfO
FCRD
Eacxxt 2 Or Sadan Hatchback L
1.9 L
L4
2 bbt
B 426
1B3BS44KUN276747
1986
CHRYSLER
D003E
Shadow
25 L
L4
1 bbt
B 428
JHMANS523GCO0O42O
1986
HONDA
HONDA
Chric Wagon
16 L
L4
2 frb*
B 436
4P3CS34TXME109053
Nft
1991
CHWaEfl
PLYMOUTH
Laiar
B 43®
JT2EL31GOH0069950
1987
TOYOTA
TOYOTA
Tareal
15 L
L4
2 bbl
B 441
1Q1AB08C5E A125660
1984
at
CHEVROLET
Chavatla CS
16 L
L4
2 bbt
B 445
JT4YR29V6G5O07140
1966
TOYOTA
TOYOTA
Cargo Van
22 L
L4
ER
B 447
1G4AM69A7CH145805
1982
GM
BUlCK
Ragal Llmttad
36 L
V6
2 bttf
B 451
1P3BP36K4HF153374
1987
CHRYSUER
PLYMOUTH
Ratlant, Custom, SE, LE
25 L
L4
1 bbl
6 452
1QMCU06D2LT200405
tto
1990
(M
PONTIAC
Transport
31 L
V6
Ft
B 460
1FABP46F7O4168B30
1983
PCfO
FCflD
TtmndartW 2 Of. Sadan
50 L
V6
2 bbt
B 466
JT2AE72EXD2096772
>933
TOYOTA
TOYOTA
Corolla
1.6 L
L4
2 bbt
B 467
1H GBA74 3XG A063308
(086
fOC*
Accord
2 0 L
14
EFI
B 460
JT2EL31D5H0075 958
1967
TOYOTA
TOYOTA
Tarcai
15 L
L4
2 bbt
B 4 72
1P38MI8C6E0304919
1984
CHRYSLER
PLYMOUTH
Horizon
22 L
L4
2 bbl
B 462
1C3CJSt£3HQ165Q66
1967
CHRYSLER
cmSLEPi
La Baron
22 I
L4 TURBO
2 t>bf
Emlaalon
Country
Comment*
Control
System*
AIR/EVPITWC/6QR(CLL
USA
CAR S»dan 4 Door 4 Window Noidtoac*
AIR/EVPaWC/EQR/CLL
KOREA
CAR
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR Hardtop
EVP>TWOEGfVCLL
USA
TRK King Cab
AJR/EVP/TWC/EGR/CLL
CANADA
CAR Station Wagon 4 Door 2 $aat
EVP/CAT/EQR/
JAPAN
CAR 4dr Sedan
A1R/EVP/TWC/EGR/CLL
USA
CAR Coupa 2 Door Notchbaok Spatial
AIR/EVP/TWOEGR/CLL
JAPAN
CAR 4 Or. Sadan
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4 Dr. Wagon
AIR/EVP/TWC;EGR/CLL
JAPAN
CAR 2dr Sadan
EVP/TWC/EGR/CLL
JAPAN
CAR 4*4 Wagon
AIR/EVP/TWC/EGR/CLL
USA
CAR Coupe 2 Door Notchbaefc
EVP/TWQEGR/CLL
JAPAN
TRK Short Bad
AfR/EVP/TWOEGR/CLL
USA
CAR Coupe 2 Door Nottffcac* SpadaJ
AtR/EVP/CAT/EGR/
USA
CAR Coupe 2 Door Nofchbae* Special
AIR/EVP/TWC'EGR/CLL
USA
CAR Sadan 2 Door Hatchback
A1R/EVP/TWC/EGR/
USA
CAR
EVP/TWC/EGft«CLL
U.S.A
CAR 2 Or. Hatchback
EVP/TWOtGfi/CLL
JAPAN
CAR Station Wagon
CAR
AlR/EVP/TWaEGR/CLL
JAPAN
CAR 3 Dr. imbat*
AIR/EVPfTWC/EGR/CLl
USA
CAR Sedan 2 Door Hatchb«c*
AiR/EVP/TWC/EGR/CLL
JAPAN
TRKTRK
Alft/EVP/TWC/EGR/CLL
USA
CAR Sadan 4 Door Ncrtchbacfc
evp/twc/eqr/cll
USA
CAR 4 Or. Sad an
EVPfTWC/EGR/CLL
USA
MPV All Purposa Vahlda
AIR/EVP/TWC/EGR/CLL
USA
CAR
Alft/EVP/TWC/EGR/CLL
JAPAN
CAR 4 Or. Sedan
EVP/TWC/EGR/CLL
USA
CAR 4 Ooof Sedan
AIR/EVP/TWC^EGR/CLL
JAPAN
CAR 2 Or Coup* All Trac
EVP/TWC/EGR^CLL
USA
CAR 4 Dr Hatchback
EVP/TWC/EGRASLL
U.SA
CAR 2 Dr SadarVCoupe
-------
Table 4-5
Vehicles Solicited But Not Instrumented with 3-Parameter Dataloggers
Sarrpla
VIN
VIN
Model
Manufacturer
Maka
Model
Engine
Fuel
Emission Control
Courlry
Commentt
Number
Confirmed Yea/
Induction
Systems
by Decoder
$001
E140KAH321J
1870
FCJO
TOFD
E150 Cargo Van
2 3 L L4 TURBO
2 bbl
AIR/EVP/CAT/6GR/
USA
TRK Econoline Van RAT : < 6000
SOOfi
1G2WK14W0JF265519
1966
CM
PCNTIAC
Oand Prik L6
2 6 L
ve
F!
AlRfEVP/TWCfEGRfCU
USA
CAR 2 Door Coupe'Sedan
$ 010
1G1AW10W4G6109607
1966
CM
CHEVROLET
Celebrity
2 8 L
v$
Fi
AIR/EVP/TWC/EGR/CLL
USA
CAR Sedan 4 Door 6 window Notchbad*
S 01 7
1G3NT27U4G0376652
1966
CM
OLDSMD0JLE
Calais Supreme
2SL
L4
Fl
EVP/TWC/EGR/CLL
USA
CAR Coupe 2 Door Notctibacfc
$ 020
S 02 1
1G1AW27XXE6SO2201
1964
OA
Chevrolet
Celebrity
28 L
V6
2 bb
Alfij£VP/TWC/EGR/Ctl
USA
CAR Coupe 2 Door Notdibeefc
S 022
1HGCA6280JA000240
198 8
HONDA
HQWJA
Accord
2 0 L
L4
2 W
EVP/TWCCLL
USA
CAR 2dr Sedan (auto)
S 026
A6AC57A330724
1076
AMERICAN
AMERICAN
Hornet
258 CI IS
1 bbi
AIRjEVP/CATVEGR/
USA
CAR 4 Dr. S«4ot
S02S
WBA 4007548
1076
HONDA
HONDA
CVCC Wagon
15 I
L4
3 bbi
EVP/
JAPAN
CAR 5 Door
S 031
son
KL2TN5462LB317077
1090
OA
PONT1AC
L« Mam
1.6 L
U
Fl
EVP/TWC/EGfiJCiL
KOFEA
CAR 4 Door Sedan
S 034
2G2AF61R4J 9232640
1988
CM
PONTIAC
Pontiac 6000
2.5 I
L4
Fl
EVP/TWC/EGflX:iL
CANADA
CAR 4 Door Station Wagon
SO 3 5
1G4HR54C9JH426906
1966
CM
BUiCK
LeSabre Limited
3 8 L
V6
Fl
ew>/twc/egra:ll
USA
CAR 4 Door Sedan
S 030
2G3AJ51R5J9306520
1066
CM
OLDSMMHE
Cutlass Ciera
2 5 L
14
Fl
EVP/TWC/EGR/CLL
CANADA
CAR 4 Door Sedan
S 050
«W82L268701
1978
RflD
FOR?
Granada 2 Or Sedan
250 a
1 bix
AIWBVP/CAT/EGR/
USA
CAR
S 060
S 064
2FABP43F7GX114162
1986
POPD
PDPD
LTO Crown Victoria 4 Or. Sedan
5.0 L
V8
2 bb
AIR/EVP/TWC/EGR/CLL
CANADA
CAfl
S 075
1G1AZ37G0ER1T4613
1984
OA
CHEVROLET
Monte Carto S
5.0 L
V8
4 bO
AIR/EVP^TWC/EGR/CLL
USA
CAR Coupe 2 Door Notcfcbeck Sped*
S083
WAUFB0856EAC729O6
1984
AUDI
AUOI
4000 Coupe or Quattro 4 WD
2.1 L
L5
Fl
EVP.TWC/CLt
GERMWT
CAR 4 Dr. Custom
S 045
JF2AM63B7CE4&7554
1982
SUBARU
SUBARU
a
1.8 L
HA
2 bfal
Alft/EVP/TWC/EGR/
JAPAN
MPV Station Wagon
S 066
1FTCR15X1LPA46236
1990
FORD
POPD
Rsng#f Regular Cab
4.0 L
V6
EFI
USA
TRK Light Trud* Ranger HAT : 4O01 > 5000
S 067
1GNCT16Z2L8116030
1090
CM
CWEWOLET
Small Conventional Cab 4x4
4 3 L
VS
Fl
AIR/EVP/TWC/EGR'CLL
USA
MPV 1/2 ton RAT : 4001 - 5000 Ibe
$ 091
2G1WN54TXL11S6917
1 900
OA
C«E>mfT
Lumnia Eurospert
3 1 L
V6
Fl
EVP'TWCJEGR^LL
CANADA
CAR 4 Door Sadan
s 000
1G1AXC8X3CT126925
1962
CM
CHEVROLET
Citation
2 8 L
VS
2 bU
AIR/EVP/TWC/EGR/CLL
USA
CAR Sedvt 2 Door Hrtchbacfc
S 102
t FAPP9S07L W f«03 |f
1900
RflD
FCfiD
Escort 4 Sr. Sadan Ha(ct\badf LX
T9t
Li
2 btf
EVPtTWC/EGRiCLL
USA
CAR
S 104
JE3CU36X4KU069206
1689
CHRYSLER
EAGLE
Summit DL
1.6 L
U
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR Station Wagon (Jmpon)
s loe
RA32O34902
No
S 11 s
JA3CU26X3LU017646
1000
MTSUBiSMI
MITSUBISHI
Mr age
1.5 L
L4
UP\
JAPAN
CAR
S 116
JHMCB7652LC072365
1000
HONOA
HONDA
Accord
2.0 L
U
Ft
JAPAN
CAR 4dr Sedan (auto)
5 121
AL10073510
No
1060
TOYOTA
TOYOTA
CAR Sedan
S136
1FABP42E0JF199672
1088
KFD
P0FD
M»tang GT 2 Dr Sedan Hatcttaadi
5.0 L
V3
ER
AtR/EVP/TWC/EGR^CLt
USA
CAR
S 130
JM2UC2213E0816360
1084
MAZDA
tMZDA
B2000 Truck
2 2 L
L4
2 btt
E VP/T WC/EGRA2 LL
JAPAN
TRK 4 Or Sedan
S 140
JN6ND06S6GW105226
1086
NISSAN
NISSAN
Truck
2 .4 L
U
EFI
AIR/EVP/TWC/EGR/CLL
JAPAN
TRK PtlMnder
S 141
JA7FL24WOLP0J0527
1000
MITSUBISHI
MITSUBISHI
Tiuch 4*2
2.4 L
t.4
MPI
JAPAN
TRK
S 142
2HGEO634XLH5204O5
1000
H0MM
K#C*
Civic 1500 H8
1.5 L
LA
MR
EVP/TWC/CLL
CANADA
CAR 2dr Hatchback ("mar)
S 151
lFABP4637EHl$tC73
1084
FCFD
FORD
Thurtderbird 2 Dr. Sedan
3 8 L
v»
2 bbi
AIR/EVP/TWC/EGR/CLL
USA
CAR
S 155
1ZVPT20CSL5206320
1990
KPD
FORDAMZDA
Probe GL 3Dr. Sedan
2 2 L
L4
EFI
' EVP/TWC/EGR^LL
USA
CAR
S 150
1G4AL10E4ED44O542
1964
OA
BU*CK
Canlury Limited
30 L
V6
2 bbi
AIR/EVP/TWC/EGR/CLL
USA
CAR Sedan 4 Ooor Notdibacfc
S 166
F15YK8S4013
1076
PQRD
Kff)
F15C Light Truck
4 9 L
16
4 bbi
AIR/EVP/CAT/EGR/
USA
TRK Light Truck Reg Ctb RAT < 6000
S 160
JG1MS2460LK743566
1990
CM
SUZUKI
Geo Mefro
1 6 L
L4
Fl
EVWTWC^GR/CLL
JAPAN
CAR 2 Ooor Hetchbadi/Liftback
S 171
1B4 FK40K6JX292532
1966
OPVSLffl.
CCOGE
Dodge Caravan
25 L
L4
EFI
AIR/EVP/TWC/EGR/CLL
USA
lyPV
S 173
JNIHZ14S2GK141618
1986
NISSAN
NISSAN
300ZX
2 9 L
V6
Fl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 2df Coupe
S 174
6K92L160963
1976
KfO
FCFD
Maverick Sedan 4 Dr.
250 CI 16
1 bbi
AIR/EVP/CAT/EGR/
USA
CAR
S 162
S 185
JHMA05334EC067076
1984
fCNO
fCNOA
Accord
1 6 L
L4
2 bbi
Afl^EVP/CAT/TWC/EGR/CLL JAPAN
CAR 2 Dr Hatchback/Coupe
S 186
JT2AE82E3G3336773
1966
TOYOTA
TOTOTA
Corolla
15 L
L4
2 bti
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4 Dr. Sedan
S 190
SJD-3038E98
1078
HONOA
H0WA
Accord
1.6 L
L4
3 bbi
EVP /
JAPAN
CAR 3 Ooor
S 191
1FTBR10C5JUB67141
1988
KFD
POPD
Ranger Super Cab
2 0 L
L4
Fl
EVPfTWC/EGR/CLL
USA
TRK Light Truck Ranger RAT : 3001 - 4000
S 193
1N6ND11S4JC 361120
1 986
NISSAN
NISSAN
Truck
2.4 L
Li
ER
Al R/E VP/T WC/EG R/C LL
USA
TRK Re^Jiar Bed
S 18 5
JT4RN50R3G0127582
1986
TOYOTA
TOYOTA
Land Cruiser
2.3 L
Li
EFI
AIRrEVP/TWC/EGR/CLL
JAPAN
TRKTRK
S 208
JF2AN53B3JE427176
1988
SUBARU
SUBWU
GL
1.6 L
H<
Fl
E VP/ T WC/EGR/C L L
JAPAN
MPV Station Wagon
S 21 3
JM2UF111XG06S3S35
1966
MAZDA
MAZOA
B2000 Truck
22 L
L4
2 bbi
AIR/EVP/TWC/EGR/CLL
JAPAN
TRK Short Bed
S 223
lYlSK1943GZ16aS09
1986
OA
CHEVROLET
Nova
1.6 L
L4
2 bbi
AIR/EVP/TWC/EGR/CLL
USA
CAR Sedan 4 Door & window Notchbacfc
S 225
2P4FH4536LR660001
1990
OflYSUR
RVM0UTH
Plymouth Voyager
3 0 L
Vf
EFI
AIR/EVP/TWC/EGR/CLL
CANADA
MPV Wagon RAT : 04001 05000 GVWR & H
S 228
JT4 VN63COJC02066 8
1936
TOYOTA
TOYOTA
Truck 4x4 Stan Bed &h. WB
3 .0 L
V6
EFI
AIR'EVP/TWC/EGR/CLL
JAPAN
TRK TRK
S 232
JP3BA24K 7FL'10915S
1985
CmYStffl
PLYWWTS
Colt, E, DL. Pram
1.5 L
Li
2 bU
AIRr&VP/TWC/EGR/CLL
JAPAN
CAR 3 Dr. Hatchback (Import)
S 234
1G1A23798EB1OS083
1984
CM
OCflWLET
Monte Carlo &
26 L
V6
2 bU
AIR/EVP/TWC/EGR/CLL
USA
CAR C014M 2 Door NotcHiack Speed
S 237
JT2SV2IE3L34182M
1990
TOYOTA
TOYOTA
Carory
20 L
Li
EFI
EVPiTWC^EGR/CLL
JAPAN
CAR 4 Or. Sedan
S 24 5
1-B3BV41B9CG1466B 5
1 982
O-RYSLEfl
DCDGt
<00
22 L
U
2 bbi
Al R»E VP/TWC^G R t CLL
USA
CAR 2 Dr. SedanX>3upe
S 247
1G3AN69YIEX3347J3
1984
3A
0LD&M0BHE
Delta 68 Royal!
5 0 L
V8
4 bbi
Al R^E VP/T WC/t G R y C LL
USA
CAR Seder 4 Door Notc^iback
-------
Table 4-5 (Continued)
Sample
VIN
VIN
Model
Manufacturer
M*e
Model
Engine
Fuel
Emia*ien
C&ontry
Commenta
Number
Confirmed
Year
Inductf on
ConUol
by Oecoder
Syatama
6 001
1FTDF15Y4HNA67137
1997
FCFO
FCFO
F150 Lighl Truck
4 9 L
16
EFi
Alft/EVP/TWC^EGFl/CU.
USA
TRK Light Trvcfc Rm Cab FIAT : 5001 - 6000
B 012
1FABPS7U4JA2232B2
i960
FCFO
FCFO
Teurue GL 4 Dr Station Wagon
3 0 L.
V6
EF1
EVP/TWC/CiL
USA
CAR
B 015
1FAPP9194KT137977
1989
FCFO
PCFO
Eecort 2 Or. Sedan Hatchback
19 L
L4
2 bbi
EVP/TWCyEGFVCLL
USA
CAR
B018
lG3HN54CeMt012733
1901
at
otosMoatf
Oerfta 08 Royale
3 0 L
V6
Fl
EVP/TWC/EGR'CLL
USA
CAR 4 Doer Sedflri
B 020
1G4BR51Y3KA494971
1989
94
BUCK
LeSebre E«tato Wagon
50 t
V8
4 bW
E VP/TWC/EGR'CU
USA
CAR 4 Door Station Wagon
S 021
lG6CDf3a2M4277240
1991
at
CADILLAC
De VWt
49 L
V6
Fl
AIR^VP/TWC/EGF^CU
USA
CAR 2 Door Coupe Sedan
B 022
IG3CW53UM4310426
1991
Oft
OLDSMOflLE
90 Regancy Brougham
3 8 L
V0
Ft
EVP/TWC/EGFl'CLL
USA
CAR 4 Ckxx Sedan
S023
1B7FD04H6GS087280
^906
CfrftfSlER
vax£
Dodge Ram Picfc-up 4X4
3.7 L
V6
1 bbt
Alfl/EVP/TWC/EGFt/CLt
USA
TRK Corvanional Cab FIAT : 04001-05000 GWR 8
B 024
002?
286WB23TXHK2«6fS4
*987
a*T*Sl£R
tCOQE
Dodge Am Van/Wegon
52 L
V6
2 bfel
AIR/EVP/CAT/EGR/
CANADA
INC Van PIAT : 05001 -07000 GVWfi A H
B 029
1FTEF15N6JNA5905O
1980
FCFO
FOD
F150 Light Tajc*
50 L
V0
EFI
AIR/EVP/TWC/EGfVCLL
USA
TRK Light Truck Reg Cab RAT - 6001 • 7000
B 031
1G3AJ51W6HG324243
1919
Ot
0LDSM08LE
Cuttata Ci*ra
28 L
V6
Fl
EVP/TWC/EGfVCLL
USA
CAR 4 Door Sedan
B032
2MECM74f5MX«12»
1991
FCFO
«FOJW
Grand Maiquia 4Pr. S«d«i GS
50 L
V6
2 bW
AiR/EVP/TWC/EGR/CLL
CANADA
CAR
B 033
1GCC714B5E2171384
1 984
CM
CHEVROLET
Smatt Conventional Cab 4i4
28 L
V6
2 bbt
AlfVEVP/CAT/EGR/
USA
TRK 1/2 ton RAT : 4001 • 5000 fca
B 034
383BK4603KT993747
1989
CHRfSLER
procf
Ariw Cu»tooi LE
22 L
L4
Fl
EVPrtWC/EGRiCU.
kEXCO
CAR 4 Dr. Sedan
8 035
2MEBM75P6KX67S910
1989
KPD
tXTCLRV
Grand M*rqiii 4Dr. S»d«r IS
5 0 L
V8
2 bbl
A1 fVE VP/T WC/E G R/C LL
CA*4ADA
CAR
B 036
1GN071825M8146406
1991
at
CfCVftXCT
5m»J( Con^fi'oraJ Cab 4i4
4 3 L
V6
F(
Alfl/EVP/rWC/EGFVCLL
USA
MPV 1/2 ton RAT : 4001 • 5000 Ibt
B 03ft
1FABP52U6KA150932
1989
FCFO
PCFO
Tauaia GL4 Dr. Wa^cn
3 0 L
V6
EF1
EVpyTWC/EGWCLl
USA
CAR
B 041
2P4FH51G4GR751156
198$
OflVSlfR
RJf MOUTH
Plymouth t/o^agar
2 5 L
L4
2 bbt
AtFUtVP/TWCfEGR/CU.
CANADA
MPV Wagon RAT : 04001-05OOT GVWR 0 H
B 051
lFABP44ElXFi14»90
1989
PCFO
FCFO
HutUnj LX 2 Dr Sadan Ccny*n»bl»
5 0 L
V8
EFl
AJR/EVP/TWC/EGR/CU.
USA
CAR
B 054
tGWCTlftR«K0l 55353
1989
CM
CHEVflCtET
Small Cofiw«niona3 Cab 4n4
2 8 t
V8
Fl
AIR/EVP/TWC/EGR/CIL
USA
MPV 1/2 ton FIAT : 4001 - 5000 Iba
BOS?
fMTEflCMAMSEABlf
B 061
1FMCU14T6GUBU126
1986
FCFO
PCFO
Br oft co ]| 4X4
29 L
V$
EFl
EVP/TWC^GFVCa
USA
VAN BtorvCQ RAT : 4001 • 5000
B 063
1FTCR14A8GPA02256
1989
PCFO
FCFO
Ranger FWgu'ar Cab
23 L
14
EFl
EVP/TWCEGR/CLL
USA
TRK Ught Truck Rr;«r RAT . 4001 - 5000
B 064
2GCCO40SC1177157
1982
at
atvRcxn
Conventional Cab GMT400
4.1 L
U6
2 bbl
AIFVEVP/CAT/EGR/
CANADA
TRK 1/2 ton RAT : 4001 • 5000 k»
B067
fG3EV!3UMUXiet60
1991
OA
0LDSM06ft£
Torena^e Trofec
3 9 L
V6
Ft
EVP/TWC/EGffCU.
USA
CAR 2 Door Coupe Seda/>
B 068
1G2FS23 EXML205941
1991
CM
pontiac
Firabird
50 t
V8
Fl
E VP^TWC/EGfVCLL
USA
CAR 2 Doc* Hatchfeack/Liftbaefc
B 060
B 070
JN1MSMP8MW003819
?9»f
NS5AN
NISSAN
200SX 2 Dr. Sedan
1.0 L
L4
Ft
E VP/TWC/EG FVC LX
JAPAN
CAR 2dr Coupa
B 071
1FAPP128 3MW 223855
1991
P3RD
PDFO
Escort GT 2 Dr. $t*r\ Hatchback
18 I
t4
EFl
EVP/TWC/EGFVCLL
USA
CAP
B 077
lG1A£27P>E72$054f
1084
at
C*€VB0tET
CavaJier Type 10
20 L
L4
EFl
AIR/EVP/TWC/EGR/CLL
USA
CAR Coupe 2 Door Nolchbacfc.
B 079
1G1JC1111K7158004
1989
04
O€VW0t£T
Cavaliar
20 L
L4
Fl
EVP/TWC/EGFl'CLL
USA
CAR 2 Door Cog pa Sedan
BOBS
1B3BA64E2EG275310
1984
OftfSlER
DODGE
Baytorit (US)
22 L
UTURBO
2 bbl
EVP^TWC/EGFVCU.
USA
CAR 2 Dr. Hatchback
B085
1FABP29U1GG270024
1086
fa©
PCPD
Tauiua 4 Dr. Sedan
30 L
V6
ER
AlfVEVP?TWC/EGR/CLL
USA
CAA
B 093
1G1JC5116K7194911
1989
CM
0€Vrot£T
Cavalitr
20 L
L4
Fl
£VP/TWC/EG?VCLL
USA
CAR 4 Ockt Se^an
B 095
1FTBR10S3EUB42009
1984
FCFO
rcflD
Ranger Super Cab
2 8 L
V«
2 bbl
EVP/TWC/EGfVCU
USA
TRK Ught Truck Ranger RAT : 3001 • 4000
B 067
2G1WL54T4M111S073
1991
at
CF€VRa£T
lumina
31 L
V6
Fl
EVP/TWC/EGFVCU
CANADA
CAR 4 Doer Sedan
B 100
7E35F526110
1977
FCFO
uevum
Monarfc 2Dr Sedan
302 CI
V8
2 bbl
AIR/EVP,CAT/£GR/
USA
CAR
B 102
1N4GB2IS5KC782419
19S9
N6SAN
NISSAN
Sentia
16 L
L4
Fl
AIR/EVP/TWC/EGR/CLL
USA
CAR 4dr Sedan
B 1 03
CE239B669962
1971
AMEHCAN
JEEP
USA
B 111
INTERCHANGEABLE
6 1 14
fG3AMI9£4EG3f64«
1984
at
OOSMOCiLE
Ct>JiM Giera Broughajn
30 L
V8
2 bbl
AI(VEVP/TWC52
1991
CWS4£R
0C0GE
Oodfte Caravan
25 t
i4
EFJ
AIR/EVP/TWC/EGR/CLL
CANADA
MPV Wagon RAT 050C t -06000 GVWR 0 H
B 1 21
YV1AX604XH1244342
7987
VCLVO
vctvo
240 Sedan 0 Wagon
23 L
14
Ff
evprrwc/CLi
CAR 4 door a
B 1 22
JT2AL22G2B2222699
1981
TOYOTA
TOYOTA
Tercel
1.5 L
I*
2 bbl
AJR/EVP/TWC/EGR;'CLL
JAPAN
CAR 3 Or. Liftbeck
B1 25
1FAPP36XOUK158O01
1991
FCFO
por>
Tempo 4 Or. Sedan GL
2 3 L
L4
CFI
EVP/TWC/CGFVCLL
USA
CAR
B 1 27
1G1JC54G6M J 1 94551
1991
04
CFCVWXHT
Cavatiar
2 2 L
L4
TBI
AIR/EVP/TWC/EGR'CLL
USA
CAR 4 DoctSedar
B 1 28
1FMCUUT5KUA42656
1959
FCFO
POD
Bronco li 4X4
2 9 L
V6
EFl
EVP/TWC/EG^CLL
USA
VAN Br ooCO RAT : 4001 - 5000
B 1 29
JT4RM67PXH504B277
1967
TOYOTA
TOYOTA
TrvcA 4x4 SB Ex»a Cab LWB
2 3 L
14
EFl
AIR/EVP/TWC'EGFVGLL
JAPAN
TRK TR(
B 1 31
F25MEZ04095
1978
PCFO
PCFO
F250 bsh> Tnpck
50 I
VI
4 bW
A IFl/EVP/C AT/EGR/
USA
TflK Ught Truck Peg Cab RAT • 6001-8500
B 1 32
1G1GZ37G2GR110555
1986
at
CHEVROLET
Monte Carlo S
5 0 L
V8
4 bb
AIFVEVP/TWC^EGR/CLL
USA
CAR Coupa 2 Door Notchback Speoa)
B 133
1GUZ3799DB142009
1983
ou
OEVflOUTT
Monte Ca/lo S
3 6 L
V6
2 bbf
AIFVEVP/TWC/EGR/CLL
USA
CAR Coupe 2 Ooor Notc^back Speaal
B 134
1G3AK47Y7EM45O270
1994
Ot
OLDSMC0LE
CuH*m CaWa
50 L
VB
4 bbl
AIR/El/P/TWC/EGR/CLL
USA
CAR Coupe 2 Door Nolchbaek Spedal
B 135
1G1TB65C9FA154310
1985
QJ
acvncuT
Chevette CS
16 I
L4
2 bbl
AIFVE V P/T WC/ EG R/CLL
USA
CAR Sedan 4 Door 6 Window P1a;n Back Hatchbacfc
B 1 36
CLNl4A0241O66
1980
at
CHEVROLET
Luv
IB L
L4
Z btt
AtR/EVP/CAT/EGR/
USA
TRK 1/2 ten
B 1 41
1B3BZ45C6GD222454
1986
OflYSLER
DCDGE
ChaiQai
22 t
L4
2 bW
EVP/TWC/tGR'CLL
USA
CAR 4 Or. Hatchback
B 146
J T 2ST65C5H7 1 06008
1987
TOYOTA
TOVOTA
Celica
20 L
L4
EFl
EVP/TWC/EGR/CU.
JAPAN
CAR 2 Dr. Cotpe
B 1 47
1GS0W51Y0J97132M
1986
on
CADILLAC
Brougham
50 t
V8
FJ
AIFVEVP/TWC/EGR/CLL
USA
CAR 4 Door Sedan
-------
Table 4-5 (Continued)
Sample
VIN
VIN
Model
Manufacture*
Make
Model
Engine
Fuel
Emiaaion
Country
Commanta
Number
Confirmed
Yttf
Induction
Control
by Decoder
Systems
8 146
1FABP40A1HF150461
1967
R3D
POFD
Mustang LX 2 Or. Sedan
2 3 L
L4
1 bbl
EVP/TWC/EGftCLL
USA
CAR
B 140
JHMSM5429BC042513
1 961
HOMM
K>OA
Accord
1.7 L
L4
2 bbl
EVP/CAT/EGR/
JAPAN
CAR 4dr Sadan
B 151
J9F17NH064652
1979
JSP
JEB>
Cherokee. MPV, 2 Or Wagon/Wt
304 CI
V6
2 bbl
AIR/EVP/CAT/EGR/
USA
TRKTRK RAT: 6025
B 152
1HGCA6281KA019512
1969
HOT*
H0K)A
Accord
20 L
L4
EFI
EVP/TWC/CLL
USA
CAR 2dr Sadan (auto)
6 155
1FABP26M2F1553600
No
1965
R7D
PCFO
Mustang II 3Dr. Hatchback
5 0 L
V6
CFI
AIR/EVP/TWC/EGR/CLL
USA
CAR
8 150
1G3AJ51W8KG342307
I960
out
OtDSMOBLf
Cudaaa Ciara
2 6 L
V6
Fl
EVP/TWC/EGftCLL
USA
CAR 4 Door Sadan
8 160
2B4GK2536MR172629
1991
OfftSlER
0CDGC
Dodg* Caravan
3 0 L
V6
EFI
AIR/EVP/TWC/EGR/CLL
CANADA
MPV Wagon RAT : 05001-06000 GVWR 4
B 161
1GTDK14K6KE527575
1 989
Ot
GM)
Conventional Cab GMT400 4x4
5.7 L
V6
Fl
AIR/EVP/TWC/EGR/CLL
USA
TRK 1/2 ton RAT : 5001 • 6000 ba
8 162
1MEPM6040MH619227
1991
PCFO
Cougar 2 Or. Sedan LS
3 6 L
V6
EFI
EVP/TWC/EGR/CLL
USA
CAR
B 166
1P3BP36DXHF290238
1967
OfttSLER
PLYMOUTH
Refianl, Custom. SE. LE
2 2 L
L4
Fl
E VP/TWC/EG R/C LL
USA
CAR 4 Or. Sadan
8 166
1C3BF66P5FX553570
1965
OftfSlER
owslb*
Fihh Avenue
5 2 L
V8
2 bbl
AIR/EVP/TWC/EGR/CLL
USA
CAR 4 Or. Sadan
8170
WVWEA016XJW3115S3
1968
VW
VW
Jatta
1.8 L
L4
Fl
EVP/TWC/CLL
GEFMANf
CAR 4dr Basic
8 171
1FMOU34X1MUC95406
1991
FCFO
PCFO
Explorer XL
4.0 L
V6
Fl
EVP/TWC/EGFVCLL
USA
VAN 4x4 Wheal Drive
B 172
JT2RN61R5K5006631
1969
TOYOTA
TcnroiA
4i4 Truck Short Bad
2.3 L
L4
EFI
AIR/EVP/TWC/EGR/CLL
JAPAN
TRK
8 174
CCU148B156579
1978
Ol
OEVTOJET
Conventional Cab
305 CI
Vfl
2 bbl
EVP/CAT/EGR/
USA
TRK 1/2 ton
8 175
1HGED3659J AO26863
1968
HOT*
OA
Civic 1500 Sadan
15 L
L4
2 bbl
EVP/TWC/CLL
USA
CAR 4dr Sadan (auto)
8 176
1P3BM44C3E0293272
1964
a#*Sl£R
PLYMOUTH
Turiamo
2 2 L
L4
2 bbl
EVP/TWC/EGR/CLL
USA
CAR 2 Dr Hatchback
8 177
YV1FX6846H2186523
1987
vavo
VOLVO
740 Sadan 6 Wagon
2 3 L
L4
Fl
EVP/TWC/CLL
SWEDEN
CAR 4 doora
B 170
1G1BN51E6KA131064
1989
on
CHEVROLET
Caprice Classic
5 0 L
V6
Fl
EVP/TWC/EGR'CLL
USA
CAR 4 Door Sadan
8 180
1G3WS14WSKD345586
1969
GM
0LD5M06ILE
Cudaaa Supreme SI
2 6 L
V6
Fl
EVP/TWC/EGR'CLL
USA
CAR 2 Door Coupe Sedan
8 161
1Y1SK5149JZ025621
1966
Ol
CfCVRCXET
Nova
1.6 L
14
2 bbl
AIR/EVP/TWC/EGR/CLL
USA
CAR 4 Doer Sedan
B 162
8 163
1G2NE14U6KC713016
1969
GM
PONT1AC
Grand Am LE
2 5 L
14
Ft
EVP/TWC/EGR/CLL
USA
CAR 2 Door Coupe Sadan
B 165
1GCCW60H7ER106666
1964
CM
CtCVROET
D Camrno 6 CabaNero
5 0 L
V6
4 bbl
EVP/CAT/EGR/
USA
TRK 1/2 ton El Camino Ca RAT : 4001 - 5000 Ibe
8 166
1G1JC1117H7159408
1967
GM
Cf€VR0LET
Cavaliar
20 L
L4
Fl
EVP/TWC/EGR/CLL
USA
CAR 2 Door Coupe^edan
B 200
1G4JS5116KJ409647
1969
Ol
BUCK
Skyhawk
20 L
L4
Fl
EVP/TWC/EGFVCU.
USA
CAR 4 Door Sedan
6 202
JM1BD2310E0760O14
1964
MfcZDA
MAZDA
GLC Station Wagon
1.5 L
L4
2 bbl
AIR/EVP/TWC/EGR/CIL
JAPAN
CAR 3 Or Hatchback
B 209
1G3AM19EX00420084
1963
GM
ODSM08&E
Cutfaaa Ciara Brougham
3.0 L
V6
2 bbl
A1R/E VP/T WC/EGR/C LL
USA
CAR Sadan 4 Door Notchbacfc
B 204
1G4NM27H7FM478057
1965
Ol
BUCK
Sonaraal Regal Limited
5.0 L
V8
4 bbl
AIR/EVP/TWC/EGR/CLL
USA
CAR Coupe 2 Door Notehbadi
B 205
1G3NT 54L4JM259390
1968
CM
0UBM0BHE
Cutfaaa Calais SL
30 L
V6
Fl
EVP/TWC/EGR/CLL
USA
CAR 4 Door Sedan
B 209
8 215
6047S6E661372
1976
OA
CADILLAC
Da ViR*
425 CI
V6
4 bbl
EVP/CAT/EGR/
USA
CAR Coupe 2 Door Hardtop
B 225
YV1FX8848G2051945
1986
VOLVO
vavo
740 Sadan 6 Wagon
2 3 L
L4
Fl
EVP/TWC/CLL
SWEDEN
CAR 4 doora
B 226
JF1AC4SB7HC202S56
1987
SUBARU
SUBARU
GL 10 Turbo
18 L
H4
2 bbl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4 Or S«dan
B 220
1HGE03553JA033488
1968
fOO
HONDA
Crvw 1500 Sadan
1.5 L
L4
2 bbl
EVP/TWC/CLL
USA
CAR 4dr Sedan (man)
8231
JT2AE86S9E0049002
1964
TOYOTA
TcnrorA
Corolla
15 L
L4
2 bbl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 2 Dr. Coupe
B 234
1G4EZ13L2M4404077
1991
Ol
BUCK
Riviera Luxury
3 8 L
V6
Fl
EVP/TWC/CLL
USA
CAR 2 Door Coupe Sedan
B 237
JM1FC3314J0610704
1988
MAZDA
MAZDA
RX7
13 L
ROT
Fl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR Coupe
8 230
2FABP74F1KX106679
1989
R3RD
PCFO
LTD Crmm Victoria LX 4 Or. Sadan
5.0 L
V8
2 bbl
AIR/EVP/TWC/EGR/CLL
CANADA
CAR
8 240
JN1 HS34P7KW012610
1969
MSSAN
MSSAN
200SX
2.0 L
V6
Fl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 2dr Coupe
8242
1G2FS67S0FN215032
1965
Ol
PONTIAC
Firabird
2.6 L
V6
MR
AIR/EVP/TWC/EGR/CLL
USA
CAR Coupe 2 Door Plan Back Spectd
B 244
KMHLF21J7HU166053
1987
HYUCAI
HYUNDAI
Excel
15 L
L4
Fl
AIR/EVP/TWC/EGR/CLL
KOREA
CAR
8 245
KMHLD11J6HU043637
1987
HftJNDAi
HYUNDAI
Excel
15 L
L4
Fl
AIR/EVPAWC/EGR/CLL
KOREA
CAR
B 246
IGNCT16Z0K016S229
1989
GUI
CICVRQlfT
Small Conventional Cab 4x4
4 3 L
V6
Fl
AIR/EVP/TWC/EGR/CLL
USA
MPV 1/2 Ion RAT : 4001 - 5000 Iba
B 250
1FABP132XDT137948
1983
FORD
FCPD
Escort 4 Or. Sadan Hatchback
1.6 L
L4
2 bbl
AIR/EVP/TWC/EGR/CLL
USA
CAR
B 251
1LNBU81F7JY847577
1968
FOR)
UNCOLN
To** Car 4 Dr. Sadan
50 L
V6
2 bbl
AIR/EVP/TWC/EGR/CLL
USA
CAR
B 252
1HGE03546KA028309
1969
HON*
H0TCA
Civic 1500 Sadan
1.5 L
L4
2 bbl
EVP/TWC/CLL
USA
CAR 4d Sedan (man)
8 253
JT4RN50R2H5105660
1967
TOYOTA
TOVOTA
Land Cruteer
2.3 L
L4
er
AIR/EVP/TWC/EGR/CLL
JAPAN
TRKTRK
B 259
1G1AW19R9D6815769
1963
Ol
0€VKX£T
Calabrity
2 5 L
L4
En
AIR/EVP/TWC/EGR/CLL
USA
CAR Sedan 4 Door Notchbacfc
B 261
JTflUFI 1E3M0066026
1991
TOYOTA
LEXUB
Lexue Luxury Sadan
4 0 L
V8
Fl
EVP/TWC/EGR/CLL
JAPAN
CAR 4dr Sedan
B 262
YV1GA694XJ0054177
1986
VOLVO
vavo
760 Sadan 6 Wagon
2 6 L
V6
Fl
EVP/TWC/CLL
SWEDEN
CAR 4 doora
B 264
2G2AF27X1E1219504
1964
OA
PONTIAC
Fiaro SE Coupe
2 6 L
V6
2 bbl
AIR/EVP/TWC/EGR/CLL
CANADA
CAR Coupe 2 Door Notchback
B 267
0G67F159642
1980
rcro
FCFO
Thunderbird HT 2Dr
5 0 L
V6
2 bbi
AIR/EVP/EGR/
USA
CAR
B 271
JN1GB21SOKU532349
1969
MSSAN
NSSAN
Santra
16 L
L4
Fl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4dr Sedan
B 272
YV1AX4951C1361436
1962
VOLVO
vavo
240 (w/ 3 Pt Saat Bait)
2 1 L
L4
Fl
EVP/TWC/
SWEDEN
CAR 5 doora
8 274
2TIAE94A7MC084133
1991
TOYOTA
TWOTA
Corolla
16 L
L4
EFI
AIR/EVP/TWC/EGR/CLL
JAPAN
CAP 2dr Sedan
8 276
KMHLF31J8HU117832
1987
HYltOAl
HKJNOM
Excal
1.5 L
L4
Fl
AIR/EVP/TWC/EGR/CLL
M3REA
CAR
B 277
1B3BK46K3KC422625
1989
OfnSLER
DODGE
Aria* Custom LE
2 5 L
L4
1 bbl
EVP/TWC/EGR/CLL
USA
CAR 4 Or Sedan
B 276
JNtHU11 S0GT124004
1966
MSSAN
NISSAN
Maxima
2.9 L
V6
Fl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4dr Sedan
B 279
WBADK830109206190
1983
MN
BMH
528a/533i
3 4 L
L6
Fl
EVP/TWC/CLL
GERAWf
CAR 52fte/S33i Sedan
B 263
2XMJP557XJA020762
1986
AMEHCAN
AMER€AGl£
Eagla 30 (SX/4) Pramiar
3.0 L
V6
MPI
AIR/EVP/TWC/EGR/CLL
CANADA
CAR 4 Or. Sedan
B 264
YV1FXB848K2314251
1969
VOLVO
vavo
740 Sadan 8 Wagon
2 3 L
L4
Fl
EVP/TWC/CLL
SWEDEN
CAR 4 doora
B 285
JT2MX62E1B0022724
1961
TOYOTA
TOYOTA
Cressida
2 6 L
L6
EFt
EVP/TWC/EGR/CLL
JAPAN
CAR 4 Or. Sedan
8 292
1G3BY69Y9FY342666
1965
Ol
OLDSMOGHf
Oalta 86 RoyaJa Brougham LS
5 0 L
V8
4 bbi
AIR/EVP/TWC/EGR/CLL
USA
CAR Sedan 4 Door 4 Window Notfhtacfc
B 296
1G1LW14T6LY1 25235
1990
Ol
C(€VR0t£T
Beretta
31 L
V6
Fl
EVP/TWC/EGR/CLL
USA
CAR 2 Door Coupe Sedan
-------
Table 4-5 (Continued)
Sample
VIN
VIN
Modal
Manufacturer
MA*
Model
Engina
Fual
Emiaaion
Country
Commanta
Number
Confirmed
Yaer
Induction
Control
by Decoder
Syatam*
B 300
CGU157U196422
1977
GM
CHEVROLET
Chwy Van Sport/Van Vandma
30! CI
V«
2 bW
EVP/CAT/EGR/
USA
TRK 1/2 ton
B 303
1 GIJC&111H7170046
1967
CM
CHEVROLET
Cavalier
20 L
L4
Ff
EVP/TWC/EGR/CLL
USA
CAR 4 Door Sadan
B 304
IG2HZS4C9HW322790
1987
CM
PONTIAC
Bonnavrfla LE
1.6 L
L4
2 bbl
AIR/EVP/TWC/EGR/CLL
USA
CAR 4 Door Sedan
B 90S
CCL440AI 627*9
1979
CM
CHEVROLET
Convcnlcntl Cab
57 L
V®
4 bW
AIA/EVP/CAT/E6R/
USA
TRK 1/2 ton w/ Hd Suap
B 306
1MEBP6545DW644243
1963
FCR)
fcBCLflY
Lyni 50i. Hatchback Sad an
1.6 L
u
2 bfai
AIR/EVP/CAT/EGR/
USA
CAR
B 306
1B3B049D2FF30C621
1985
CHRYSiEF
CXDGE
Ariaa Cuatom LE
2 2 L
L4
Fl
EVP/TWC/EGR/CLL
USA
CAR Station Wagon (Import)
B 309
1FAPP2699JW209976
1986
FCHD
FCR)
Eacort 4 Or. Scaban Wagon LX, GL
1.9 L
L4
2 bW
£VP/TWC'EGfVCLL
USA
CAR
B 310
1C3BH58E 4HN460772
1967
0«tSi£R
ORYSiSl
La 3a; on, GTS
2.2 L
LA TURBO
2 bW
EVP/TWC/EGR/CLL
USA
CAR 4 Or Hatchback
e 311
tN6ND11Y2GC446136
1986
NISSAN
NtSSAN
Truck
2.4 L
L4
EF1
AIR/EVP/TWC/EGR/CLL
USA
TRK Regular Bed
B 312
P10GE201518
1976
FORD
FCRD
F100 Light Truck
5 6 L
V6
2 bbl
AIR/EVP/CAT/EGR/
USA
TRK Light Truck Rag Cab RAT : < 6000
B 313
1G4AJ69A6EH82I456
1964
CM
BUCK
Regal
3 6 L
V6
2 bbl
AIR/EVP/TWC/EGR/CLL
USA
CAR Sedan 4 Door Nolchba^t
B 318
1G2BL61YSJA200424
1966
OA
PONTIAC
Safari Wagon
5.0 L
V8
Fl
AIR/EVP/TWC/EGR/CLL
USA
CAR 4 Door Station Wagon
B 321
VV1AXMSSG1661639
1966
VOLVO
VOLVO
240 Sedan 6 Wagon
2.3 L
L4
Fl
EVP/TWC/CLL
SWSB*
CAR 5 door*
B 322
JN1PB12S4FUS3I495
1965
NtSSAN
NtSSAN
Santra
1.6 L
14
2 bH
AIR/EV P/T WC/EGR/CLL
JAPAN
CAR 2dr Sadan
B 335
JN1HB1IS5CU0O6229
1962
NISSAN
OATSUN
210
1.5 L
14
2 bW
EVP/CAT/EGR/
JAPAN
CAR 4dr Sadan
B 343
TE51-696464
1975
TOYOTA
TOYOTA
Corolla
1.6 L
L4
2 bW
AIR/EVP/EGR/
JAPAN
CAR 2 Or. Sedan
B 345
6Y69A624642
1976
FCRD
UNCOLN
Ma/k IV or V HT 2 Dr.
460 CI
V6
4 bW
AIR/EVP/CAT/EGR/
USA
CAR
B 346
J72WX63E600006546
1963
TOYOTA
TOYOTA
Craatida
2.6 L
L6
ER
£VP/TWC/EGft/CLL
JAPAN
CAR 4 Or. Sadan
B 347
JM1GD2222J1532651
1986
MAZQA
MAZQA
626
2 2 L
L4
Fl
EVP/TWC/EGFVCLL
J# PAN
CAR 4 Dr. Sadan
B 348
1N4P821S1HC665474
1967
NISSAN
NISSAN
Santra
16 L
L4
2 bH
AIR/EVP/TWC/EGR/CLL
USA
CAR 4dr Sadan
B 349
WTEFCHANGE/fitf
B 352
1FMCU14T7JUB25370
1966
RHI
fOO
Bronco II 4X4
2.9 L
V6
En
EVP/TWC/EGFVCLL
USA
VAN Bronco RAT : 4001 - S000
B 955
B 356
J72EU2D3J0254164
1966
TOYOTA
TOYOTA
Tarcal
1.5 L
L4
2 bbl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 2 Or. Coupe AMTiac
B 359
1HGCA5620JA059966
1986
HOTOA
HOTfM
Accord
2 0 L
Li
ER
EVP/TWC/CLL
USA
CAR 4dr Sadan (auto)
B 360
1FABP3191GT119712
1966
P3=D
KFD
Eccorl 2 Dr Sedtft Hatchback L
IB L
L4
2 bW
Alft/EVP/TWC/EGFV
USA
CAR
B 362
B 364
B 366
B 373
1P3BP26C5DFI23152
1983
CHFNSLER
PLYMOUTH
Ralianl
22 L
L4
2 bbl
AIR/EVP/TWC/EGR/CLL
USA
CAR 4 Di. Sedan
9377
1G 3AJ19R5 FG396797
1965
34
CLD5M09LE
Culaaa Ciara LS
2 5 L
L4
En
EVP/TWC^EGFVCLL
USA
CAR Sedan 4 Coor 6 Window Notdtbacfc
B3?6
1FABP55U5KA178079
1969
FCR)
RH)
Ta^ut L 4 Di Wagon
3.0 L
V6
ER
EVP/TYYC/EGR/CLL
USA
CAR
B 379
JM1GC22A4H1100650
1967
MAZDA
MAZQA
62C
2.0 L
14
Fl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4 Dr Sedan
B 360
1B3BD36K4GF152556
1986
ORrSlEP
DCDGE
Ariaa SE
2 5 L
Li
1 bbl
EVP/TWC/EGR/CLL
USA
CAR4 Or. Sedan
B 362
WBADB7401EI194390
No
1964
BMW
bm
526e/533l
2.7 L
L«
Fl
EVP/TWC/CLL
GEPMWVf
CAR 526e/53S Sadan
B 363
1FABP5347JA150060
1966
ROT
FOT
Taunt* LX 4 Or Wagon
36 I
V6
ER
AIR/EVP/TWC/EGR/CLL
USA
CAR
B 364
JT2MX63E2E0054388
1964
TOYOTA
TCWOTA
Creaaidt
3.0 L
L4
ER
EVP/TWC/EGR/CLL
JAPAN
CAR 4 Dr . Sedan
B 365
JN1MN06S1CM004616
1962
NISSAN
DATSUN
310
16 L
L4
2 bbf
EVP/CAT/EGR/
JAPAN
CAR 2dr HB Sadan
B 367
JNIPBI1SXFU629946
196S
NISSAN
NISSAN
Santra
1.6 L
L4
2 bbl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4dr Sedan
B 366
JT2RA64L000024499
1963
TOYOTA
TOYOTA
Supra
2 4 L
L4
2 Of
EVP/TWC/EGR/CLL
JAPAN
CAR Uftback
0 402
t FABP534XJA143314
1986
ROT
fCFD
Tavrut LX 4 Or. Wagon
3.6 L
V$
EFI
AIR/EV P/T WC /EGR/C U
USA
CAR
B 404
1B3CA4 4KXJG365292
1966
ORi5i£R
DCDGE
Oaytona (U.S.)
2 5 L
L4
1 bbl
EVP/TWC/EGR/CLL
USA
CAR 2 Dr. Hatchback
B 407
1G1JC1110H7162618
1987
(M
0€VTO£T
Cavaliar
20 L
L4
Fl
EVP/TWC/EGR/CLL
USA
CAR 2 Door Coupe/Sedan
B 400
014AB7S216324
1977
CHTiSlER
DCDGE
Oodga D100
225 CI
V6
2 bbl
EVP/CAT/EGR/
USA
TRK Conventional Cab RAT : 6O0CM or Laaa
B 411
JH4KA765XMCO16O70
1991
fOM
ACURA
Lagand
3 2 L
V6
Fl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR Legand 4d< S*dan 4cpeed
6 416
YS 3AK46D6L5004566
1S90
SAAB
SAAB
900 'S' Sariaa 2dr
2 0 L
L4
Fl
EVP/TWC/CLL
SWEDEN
CAR 4 Or Sedan
e 4te
JN1FU2IP4MT313M2
1991
NISSAN
NISSAN
Maxima
2 9 L
ve
Fi
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4di Sedan
6 423
JHMEE2754MS003363
1991
HON*
HOTOA
Crvie Wa^on
1.5 L
L4
2 bbl
EVP/TWC/CLL
JAPAN
CAR 4df Wpi & Wagovtn (man)
6 427
t FA 8P64 W6HH100306
1967
PCFO
FCRD
Thund»rbkd Turbo Cpa 2 Dr. Sedan
2 3 L
U TURBO
EH
AIR/EVP/TWC/EGR/CLL
USA
CAR
B 430
1FTDF15Y4MNA52700
1991
K*D
FCRD
F150 UgM Truck
4 9 L
16
ER
AIR/EVP/TWC/EGR/C LL
USA
TRK Light Truck Reg Cab RAT : 6001 - 6000
6 432
JT2AE»4A5M3439124
1991
TOYOTA
TOYOTA
Corolla Oaluia
15 L
L4
2 bbl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4 Door Sedan
fi 437
tFABP4036FG169786
1965
rh>
FOB
LTD 4 Dr Wagon Sqvira
3 8 L
ve
2 bbl
AIR/EVP/TWC/EGR/CLL
USA
CAR
6 440
JNIHB02S3CV468445
No
1982
NSSAN
DATSUN
210
1.5 L
L4
2 bbl
EVP/CAT/EGR/
JAPAN
CAR 2dr Sedan
6 442
J0TRG5102H64SS641
1087
Out
ctevnoLET
Spectum Sedan
15 L
14
Ft
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4 Door Sedan
B 444
1HGAD7423FA060632
1985
HOCA
HOTCIA
Accord
16 L
L4
2 bbl
AIR/EVP/TWC/EGR/CLL
USA
CAR 4 Door Sedan
6446
CM43N6C170061
1976
OfiYSLER
CHRYSLW
Chryalar FuH Si
400 CI
VB
4 bbl
AIR/EVP/CAT/EGR/
USA
CAR 4 Dr Sedan HT
6 446
JT2EL31D9K0349959
1969
TOTOTA
TOYOTA
Tercel
1.5 L
LA
2 bbt
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 2 Di Coupe All Ttac
B 449
KMHLD31J3HU021953
1987
HYUN3AJ
Eieal
15 L
L4
Fl
AIR/EVP/TWC/EGR/CLL
NOflEA
CAR
B 455
1G2JB2700F7567426
1965
34
PONTIAC
Sunbtrd
16 L
L4
TBI
EVP/TWC/EGfVCLL
USA
CAR Coupe 2 Door Notdtback
B 456
JT2RA63C7F6250542
<985
TOYOTA
TCWOTA
Supra
24 L
L4
2 bbl
EVP/TWC i^GfVCU.
JAPAN
CAR 2 DrCoup*
B 456
4T1SV21EOKU049447
1969
TOYOTA
TCYCTA
Camry
20 L
L4
ER
EVP/TWC/EGfVCLL
USA
CAR 4 Dr S*d«r>
fi 459
1HGAD5435EA055864
1964
Hero*
hotcm
Accord
1 6 L
L4
2 bbl
AIR/EV P/CAT/TWC/EGfl/CLL
USA
CAR 4 Doer Sedm
B 463
1FA8P52U4HG286736
1967
rcro
KFO
TauruaGL 4 Dr Wagon
3 0 L
V6
ER
EVP/TWC/EGR/CLL
USA
CAR
fl 464
1G1JC51I3JJ1764S0
1966
GM
ocvnciFr
Cavaliar
20 L
L4
Fl
EVP/TWC/EGR/CLL
USA
CAR 4 Door Sedan
-------
Table 4-5 (Continued)
Sample
VIN
VIN
Model
Manufacturer
M*e
Modal
Engine
Fuel
Emiaeion
Country
Comment*
Number
Confirmed
Year
Induction
Control
by Decoder
Syatama
B 465
1G3BN37YSFY397447
1905
am
OUSMOOIE
Oalta 00 Roy ale
5 0 I VI
4 bbi
AIR/EVP/TWC/EGR/CLL
USA
CAR Coup* 2 Door Nolehbacfc Special
B 469
JG1MR615SJKJ02494
No
1900
CM
suzuw
Sprint
16 L U
Fl
AIR/EVP/TWC/EGR/CIL
JAPAN
CAR 4 Door Hatchback
B 470
1P3BP21C4DG136010
1903
CWNSIER
R.YMCUTH
Reliant
2 2 L L4
2 bU
AIR/EVP/TWC/EGR/CLL
USA
CAR 2 Or Seda/vCoupe
B 471
1G1LV11W5JY205140
1900
an
OtVWXET
Beretta
2 0 I V6
Fl
AIR/EVP/TWC/EGR/CLL
USA
CAR 2 Door Coupe/Sedan
B 473
1FABPS5USHA160704
1987
FCFD
RCPD
Tevrue L 4 D» Wagon
3 0 L V6
en
EVP/TWC/EGR/CU.
USA
CAP
B 474
1F ABP22X4GK125291
1900
PCfQ
POPO
Ueer 4 Dr Sedan
2 3 I L4
CFI
AIR/EVP/TWC/EGR/CLL
USA
CAR
B 475
1G1ABO0C7EY142400
1904
(M
OEVRCXfT
Chevette CS
101 14
2 bbl
AlfVEVP/TWC/EGR/CLL
USA
CAR Sadan 2 Door Hatchback
B 476
9K94T231057
1979
KJFO
KHJ
Fairmont Station Wagon
3.3 L ie
1 btt
AIR/EVP/CAT/EGR/
USA
CAR
B 470
JT2AE95C1K321202O
1909
TOYOTA
TOYOTA
Corolla
1.5 L L4
2 bbl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 2 O. Coupe
B 404
1HGCA5630JAO939O6
1900
HOKM
KNA
Accord
2.0 L L4
EF1
EVP/TWC/CLL
USA
CAR 4dr Sedan (auto)
-------
had to be available and not in use by a vehicle in the field. At the beginning of the field
work in each city, most manufacturer's late model-year vehicles could be solicited
because all of the 6-parameter datalogger systems were available for instrumentation.
However, after the first week, usually only about five 6-parameter datalogger systems
were available daily. This restricted the choice of solicited vehicles that could be
instrumented on a given day.
For most vehicles, the NGK wide-range oxygen sensor was installed on the
exhaust system of the vehicle by welding a boss on the exhaust pipe between the catalytic
converter and the muffler or resonator. The boss was positioned as close as possible to
the catalytic converter end of the pipe. At the end of the test period, the affected
portion of the exhaust system behind the catalytic converter was replaced with Walker
exhaust system parts. The exhaust system work was performed by a local muffler shop.
For Spokane vehicles with sample numbers S003, S006, S007, SOW, S019, and S028, the
wide-range oxygen sensor was placed one foot behind the muffler outlet in a four-foot
long tailpipe welded to the muffler.
Table 4-3 also summarizes the solicitation and installation activities
for the 6-parameter dataloggers. A total of 85 vehicles in Spokane and 163 vehicles in
Baltimore were solicited for 6-parameter installations. Of these, 79 instrumentations
were performed. Just as for the 3-parameter systems, the installation rate for Spokane
(49%) was higher than for Baltimore (23%). Reasons for the inability to instrument 26
vehicles with 6-parameter systems were: incompatibility of the 6-parameter dataloggers
with the vehicle, vehicle diagnostic systems showing error codes indicating some sort of
vehicle problem, and malfunctions of 6-parameter datalogger systems accompanied by a
lack of an available replacement system.
Tables 4-6 and 4-7 show the decoded VINs for the 6-parameter
instrumented and noninstrumented vehicles, respectively.
4-48
-------
Table 4-6
Vehicles Instrumented with 6-Parameter Dataloggers
Sample VIH
Number
YIN
Confirmed
by Oecodar
4^
S 003
SO06
S 007
S 0t4
S Of ft
S 028
S 040
S 042
9 045
S 047
S 051
8 053
S0S&
S 063
S 070
S082
s oee
S 092
$093
S 097
S 09ft
S 107
S 106
S 110
S 125
S 127
S 129
S 131
S 133
$ 148
S ISO
S 164
S 177
S 161
S 201
$20*
S 2 tO
S 215
S217
S 220
S 236
5 243
B 005
B ooa
BOH
B 044
B 053
B 055
B 065
B 074
B 076
B 099
B 101
B 113
6 119
B 140
B 144
B 145
B 159
B 165
B 167
JH1HJ01P61T366371
1Q2AF54T9L6235376
2G1WL54RXLI 169676
304A154N3LS605881
JM1BO?265LOUt26?
1G2HX54C9L1240876
1G3AM54N2L6363462
2FAPP37X6LBI05148
1YVG D22B9L5241604
1B3XA4636LF914626
JMtGD222XL1S22669
IFACP52U5LG167445
1FACP50UXLG149557
J T2SV21E2I0 336391
JA3CR46V8LZQ 36160
1MECM50U6LG6497U
1B4FK44R0LX295520
1P3 XA46K8LF 615124
2G1WN54T4L9216037
2B4FK4530LR632793
1G4CW54C6L1624667
1B3XA4639LF836612
JT2SV24E2L34MSS3
1N4GB22B6LC764253
1LNLM9846LY 702663
1Q3AJ54N1L6343342
1YVGD2286L5224064
IMECM5344LG63Q501
1G2NES4D1LC373695
2GIWL54T8L1197259
1F ACPS2U XLG233424
IG3HY54C9LH309730
1GUW14T8LE150508
3G4 AH54N7LS609 966
1G3HY54C9U816555
2FAPP3«K2l8t67744
JMtBG2261L0t42O55
1N4QB22BXLC764143
fG2WJS4T3LF263419
1MEPM36X2LK63I696
1G1LVI4T1LE144679
2G1WN54T9L920I394
3B3XA4633MT034156
1G1L764W8KY21B187
IB3XG24K5KG180041
1G2H254C2KW215880
1G3NT14DXKM2 55041
JM 1GD2224L1610193
1FAPP6046KH126726
1MEPMS042LHS79360
1F ACP57U7MAI75592
tG3HY14C3KW311671
1G3CW54C6K4300538
1N4G62253KC735159
1GUM4T4M7133492
1MEPM36X2KK620959
JE3CU36X6KU035123
1G3NT14 D2KM235737
IG2HXS4CXKW259695
2B4GK55R7MR201092
1YVGD22B0M5125967
Mode*
Manufacturar
Make
Modal
Engine
Fuel
Emlaelon
Country
Year
induction
Control
Sy sterna
1090
NISSAN
NISSAN
Muime
2 9 L
V6
Fl
Al fVE V P/T W C/E G R/C L L
JAPAN
1990
OA
PONT1AC
Pcrv«ac 6000 IE
3 L
V6
Fl
EVPfTWC/EGfVCLL
USA.
1990
o 4 Dr Sedan GL
2 3 L
L4
en
EVP/TWC/EGFVCLL
CANADA
1990
MAZDA
MAZDA
Proiaga
16 L
L4
EFI
JAPAN
1990
NISSAN
NISSAN
Santra
16 L
L4
Fl
USA.
1990
OA
POND AC
Grarwl Prta
3.1 L
V6
Fl
EVPiTWOEGR/CLL
USA
1990
FCFO
UEPOJR*
Topaz 4 Or. Sedan GS 4 WOR
2 3 L
L4
CR
EVP/TWOEGJVCLL
USA
1990
OA
CHEVROLET
BereHa
31 L
U6
Fl
EVPH WCEGR/CLL
USA.
1990
OA
CHEVROLET
Lumlna Euroaport
3.1 L
V6
Fl
EVP^rWC/EGR/CLL
CANADA
1991
CHRYSLER
DCOGE
Spirit
3.0 L
V6
TBI
EVP/T WC/E G R/CLL
MEXICO
I960
OA
CHEVROLET
Coralca
2 6 L
V6
Fl
evp/twoegra:ll
USA.
1969
CHRYSLER
DOOGE
Oayiona (U S )
2 5 L
L4
1 bbl
evp/twc«gra:ll
USA
1969
OA
PONT1AC
Bonneville SE
36 1
V6
Ft
evp/twoegr^ll
U S A.
1989
CM
OLDSMOBJIE
Cutiatt CalaJa SL
2.3 L
L4
Fl
evp/twocll
USA
1990
MAZDA
MAZOA
626
22 L
L4
Fl
EVP/T WCJEGRXLL
japan
1969
PCPD
PCPD
ThumJerHrd Std. 2 Dr. Sedan
3 6 L
V6
ER
EVP/TWCEGR/CLL
USA
1990
PCPD
Cougar 2 Dr. Sedan LS
38 L
V6
EFI
EVP/TWCEQRjCLI
USA
1991
PCPD
P3FD
Taurua GL 4 Dr Station Wagon
3 0 L
V6
ER
EVP/TWC-EGR/CU
USA.
1969
a*
OLDSMOGJLE
Delta 86 Royate Brougham
3 6 L
V6
FJ
EVP/TWCEOR/CLt
USA
1969
CM
OLDSM06ILE
96 Ragency Brou^iam
38 L
V6
Fl
EVP/TWC^EGR/CLL
USA.
(989
NISSAN
NISSAN
Santra
1.6 L
14
F1
AlfVEVPHWC/EGR^CLL
USA
1991
OA
CHEVROLET
Cavalier 224
3.1 L
V6
Fl
evp/twc^gra:ll
U.S A
1989
FORD
mseur*
Topaz 4 Dr Sedan GS 4 WDR
2 3 L
L4
CR
EVP/7W&EGR/CLL
USA.
1989
ORYSLER
EAOJE
Summit DL
16 L
L4
MPI
AIR/EVP/TWC/EGR/CLL
JAPAN
1989
CM
OLDSMOBH£
Cutlass Calais SL
23 L
L4
Fl
EVP/TWC/CLL
U.S A.
1989
OA
PONTIAC
BonnevIKe LE
3 6 L
V6
Fl
EVP/TWCi€Gfti«CLL
USA
1991
CHRYSLER
DOOGC
Dodge Caravan LE
3 3 L
V6
ER
AIR'EVP/TWC/EGR'CLL
CANADA
1991
MAZDA
MAZOA
626
2.2 L
L4
Fl
EVP/T WC/EGfVCLL
USA
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAP
CAR
CAR
CAR
CAfl
MPV
CAR
CAR
M?V
CAR
CAR
CAR
CAR
OA
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CM
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
CAR
MPV
CAR
RAT : 04001-05000 GVWR A H
RAT 04001-05000 GVWR & H
4dr Sedan
4 Door Sedan
4 Door Sedan
4 Door Sedan
4 I> Sedan
4 Door Sedan
4 Door Sedan
4 Dr. Sedan
4 Dr. Sedan
4 Dr. Sedan
4 Dr. Sedan
Extended Wapon
4 Dr. Sedan
4 Door Sedan
Wagon
4 Door Sedan
4 Dr. Sedan
4 Dr. Sedan
2dr Sedan
4 Door Sedan
4 Dr. Sedan
4 Door Sedan
4 Door Sedan
4 Door Sedan
1 Door Coup* Sadan
4 Door Sedan
4 Door Sedan
4 Dr. Sedan
Sdr Sedan
4 Door Sedan
2 Door Coupe Sedan
4 Door Sedan
4 Dr. Sedan
4 Door Hatcttiadi/Llftbick
2 Or Hatchback
4 Door Sedan
2 Door Coupe Sedan
4 Dr. Sedan
2 Door Coupe Sadan
4 Door Sedan
2tik Sedan
2 Door Coupe Sedan
Station Wagon (Import)
2 Door Coupe Sedan
4 Door Sedan
Wagon RAT : 05001-06000 GVWR A H
4 Dr. Sedan
-------
Table 4-6 (Continued)
Sample VlN
Number
VIM
Confirmed
fry Decoder
G 198
B 236
6 261
B 262
B 201
B 302
B 324
B 354
B 356
B 363
B 370
B 390
B 392
B 414
B 431
B 431
B 435
B 454
2«*
CANADA CAR
USA CAR
USA CAR
USA CAR 4 DOW Sedan
USA CAR
USA CAR
U SA CAR 4 Door Sedan
USA CAR 2 Door Coup* Sedan
JAPAN CAR 40r Sedan
USA CAR 4 Dr. Sedan
CANADA CAR 4 Door Sedan
USA CAR 4 Door Sedan
USA CAR4 Door Sedan
USA CAR4 Door Sedan
JAPAN CAR 3 Dr. Hatctrbads (Import)
CANADA CAR 2 Door Coupe Sedan
-------
Table 4-7
Vehicles Solicited But Not Instrumented with 6-Parameter Dataloggers
Sample
VIN
VIN
Model
Manufacturer
Maka
Model
Engine
Fuel
EmiMion Conlrol
Country
Corn menu
Nurrfcer
Cor)fnm*(t Viir
Induction
Syllemt
by Decoder
S 012
tP3XASe3SLF774797
1990
CH?YSL£R
fVYMOLTTH
Acclaim LE
3 0 L
V6
T9I
EVP/TWC/EGRCLL
USA
CAR 4 Dr Sedan
son
IG4*V54UOLM07977«
1990
(M
BUCK
Skylark
2 5 L
L4
Fl
EVP/TWC«GR«LL
USA
CAR 4 Door Sedan
S01C
JMIGD2226L1112401
1990
MAZDA
MAZDA
<26
2 2 L
L4
Fl
EVP/TWC/EGRrfUL
JAPAN
CAR 4 Dr Sedan
S 023
JMlGD2??7LtaOM76
1990
MAZDA
MAZDA
<26
2 2 L
L4
Fl
EVP/TWCEGR/CLL
JAPAN
CAR 4 Dr Sedan
S 033
1C3XJ41K1LG43I532
1990
CHIVSLER
CHRYSLER
U Garon
25L
14
1 tibi
evpawc1
JAPAN
CAR
S0!>7
lB3XCS639LD7tl7i0
1990
CHRYSLER
Q00GE
Dynaaty LE (U S)
30 L
V6
TBI
EVP/TWC/EGFICLL
USA
CAR 4 Or Sedjn
$062
1G1LT54GSL Y249993
1990
CM
CHEVROlfT
Corsica
22 L
14
TBI
AIR^EVP/TWC/EGR/CLL
USA
CAR 4 Doo> Sedan
S065
JT2AE62E4J3034661
1901
TOYOTA
TOYOTA
Corolla
1 5 L
14
2 bbl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4 Or Sedan
S068
JT3VN39WXL0O23696
1990
TOYOTA
TOYOTA
Van 4k4
30 L
V6
EFI
AIR/EVP/TWC/EGR/CLL
JAPAN
MPVMPV
$071
AG3XC55RSL0e»545
CHRYSLER
Haw tcrVir Landau
221 1.4 TVJRftO
fh
MftitVPrtWCJfcGftJCtt.
USA
CAR4f> Sedan
S 074
UNCM92EXLY7O3O10
1990
FCfiD
UNGOLN
Maris VII aau 2 Dr Coupa
SOL
V*
EFI
AIRfEVP/TWC/EGR/CLL
USA
CAR
S 077
2FACP73F4LX14B931
1990
PCflD
FORD
LTD Crown Victoria 4 Dr. Sedan
5 0 L
va
2 btt
AIRfEVP/TWC/EGR/CLL
CANADA
CAR
S060
1P4FH44ROLX303404
1990
cnw&m
RYM0U7H
Rynvulh Voyager
3 3 L
V6
en
USA
MPV Exlended Wagon RAT 04001 05000 GVWR & H
son
1FTCR14X1LPA46236
No
1990
PCflD
ROAD
Taurus
CAR
8 004
1B3XP49K1IN1665«1
1990
OflYSLER
DCOGE
Shadow
25 L
L4
1 btf
EVP/TWC/EGR^CLL
USA
CAR 4 Dr Hak^bach
soes
1G3WS54T9LD395199
1990
CM
OLDSM03LE
Ccltass Supreme SL
3 1 L
vs
Fl
EVP/TWC/EGRCLL
USA
CAR 4 Door Sedan
S 101
1FAPR»S»XIW1«0029
1990
R3RD
FCRD
Escort 4 Dr Sadan Htfcftback LX
1 9 L
L4
2 btt
E VP/T WC/EGR^C IL
LISA
Ofl
S 111
e jO
1Q3WS&4TXL0375651
1990
CM
OUKMOOLE
Cutfasa Supreme SL
3 1 L
V6
Fl
EVP/TWC/EGRjCIL
USA
CAR 4 Door Sedan
a TiJ
S 125
1G3Vm54T6LD3Vt674
1990
CM
0LDSU09A£
Cuflass Supreme
3 I L
VC
Fl
EVPAWC/EGR^IL
USA
CAR 4 Door Sedan
S 124
JHMCS7662LC076249
1990
HONDA
HONDA
Accord
20 L
L4
Fl
JAPAN
CAR 4dr Sedan (auto)
S 14 S
1FACO52U1LG214209
No
1990
FCRD
FORD
Taurus GL 4 0* Wagon
30 L
VC
EFI
EVP/TWC/EGfl^LL
USA
CAR
$ 140
3G4AH54NXLS«13221
1990
CM
QUICK
Century Custom
3 3 L
V6
Ft
EVP/TWC/CLL
rooco
CAR 4 Ooor Sedan
S 153
JT2VV22WSL010123I
1990
TOYOTA
TOYOTA
Canvy
25 L
ve
EFI
JAPAN
CAR 4x4 Wagon
$ 157
2G1WN54T2L1150707
1990
OA
CHEVfdfT
Lunina Euroiport
3 1 L
vs
Fl
EVP/TWC/EGRCLL
CANADA
CAR 4 Door Sedan
S 111
1G3AJ54N5L6319366
1990
CM
OLD6M06M-E
Cutlass Ga«a
3 3 L
ve
Fl
EVP/TWC/CLL
USA.
CAR 4 Door Sedan
S 192
1G3WR54TKL0373223
No
1990
(M
OLDSMDOUE
Cut ass Suprama Iniw.
3 1 L
V6
Fl
EVP/TWC/EGfi/CLL
USA
CAR 4 Door Sedan
5 170
6 193
1G1LT54Q2L613426Q
1990
-------
Table 4-7 (Continued)
S«mp1. VIN
Ncrrrbaf
VIN Mod*l Manufacturer
Confirmad Y*a»
bf D9CO&*
Engirt*
Fv*\
Induction
L/l
N»
B 003
B 004
2G4WD14W7K14UO90
1669
OA
BUCK
Rtgal Umitad
20 L
V6
Fl
B 00ft
tB3BA7ftJ4KF443566
1969
of?isi£n
DCDGE
Spwil LE. ES
25 L
LI TURBO
i bbi
B 009
1G2NEUD3KC 752033
1989
Oi
POMTTAC
Gfjrd Am LE
23 L
L4
F1
a 0)3
t¥VGD22B9M5123943
1991
MAZDA
WAZQA
626
22 L
L4
Fl
B 0*4
B 025
1FABP44A9KF12870I
1989
RH)
fCFD
Mv4teftg LX 2 Or S*dan Corv*r»trf«
23 L
L4
1 bbi
B 026
2G1WN54T0L9311S77
1990
at
O€VFl0LET
Lufflin* Euro# port
31 L
V6
Fl
b 02a
2B4FK2SK7KR269211
1969
CHRrSlER
DCDGE
Dodg* OAiawtn
25 L
L4
ER
B 040
2FABP74F6KX163l«7
1969
FCFD
FOR)
LTD Cio*m Victoria LX 4 Dr. S*d«n
50 L
V6
2 bbl
B 043
1FACP57U3MAI5M*3
1991
TOO
fCBD
Taurus GL 4 Dr Station Wagon
30 L
V6
EFI
B 045
JU1BG22S2M0226959
1991
MA2DA
MAZDA
Prot«Q* 4 Df Sadan
1 6 L
L4
Efl
B 04ft
1FABP50UBKA17790?
1969
TOFC
POfD
Taurua L 4 Or Wagon
30 t
V«
en
B 047
BO&(
2G4 WB14LXM1661S26
1991
34
BUCK
R*gai Cuttam
36 L
V6
Fi
B 060
1N4GB22S6KC733973
1969
NSSAN
MSSAN
S*ntr*
16 L
L4
Fl
B073
1 UEBMS5U2KA601529
1989
Pdc
VBBJH
$*bt* 40 Slrt Wp GS
30 L
V6
EFI
Boao
1LN8M63FXKV727917
1969
FCFD
UNCOLN
Town C*r Car%«r 4 Or S«d*r
50 I
va
2 bW
B062
1G1AW»1W6X6214925
1969
CM
C^CVKXfT
€0ltb4ily
20 L
V6
Fl
B 047
tG3HN54C6MM574«
1969
OA
cnnsuooLB
Mti 66 Royafc
36 L
V«
Ft
B 08ft
1G2HXMC4L1219971
1960
at
POMTIAC
Bwvillt LE
36 L
V6
Fl
e o9o
1G3WH14TXU0306045
1961
CM
ousMoaif
CuflaM Supfwna
31 L
V6
Fl
6 Oft J
1G4HPS4C1MH431160
1991
CM
BLftCK
LtSibt CuilM
36 L
V6
Fl
e o»2
1G3HN54C7HH33&477
1991
cm
odsmobie
D«tti U RoyaU
36 L
V6
Fl
B 094
1H4QB2lS*KC7339M
1969
NSSAN
NSSAN
S*ntr«
16 L
L4
Fl
e loe
1GltT53T4WVf 10010
1991
CM
0€VTOI£T
Cotcie*
31 t
V6
Fl
6 t 07
(G2WH54T9MF200743
1991
OA
powtwc
Grvtd Pir. LE
31 L
V6
Fl
B 106
1Q4ALS1 Nl KT47936S
1969
a4
BUCK
Ctntwy Umi!»d
33 I
V6
Fl
B 10ft
1G2HX44C9KW20ft714
1969
CM
POWTWC
LE
36 L
V6
f\
B 112
1G4MC54N0KM036312
1960
CM
BUtCK
Skylark Ctratom (4-OR)
33 L
V6
Fl
B 11ft
1GtJCIt1 KKil 957S1
1969
CM
CHEVROLET
Cavaliar
20 L
L4
Ft
B 1 3ft
1LMCU6lWaMY714ft2ft
1961
R3U
UNCOLN
Town Car 4 Or. S*4an
4 6 L
V»
EFI
a t42
If ABP32U2KA1597S3
1969
fcto
fCPD
Tauni«GL4 0e Wigcn
30 L
V6
ER
B 154
1G3WHS4T 6UD3260O6
1991
CM
OLOSMO0KE
CuitM Sup*am*
31 L
V6
Ft
B 157
B 173
B 107
\ G3HN&4C9UH306961
1991
at
0USM0OLE
Datta $8 Royal*
35 L
Vft
Ft
1G4ALS1N1 KT4ftft405
1969
a*
BUCK
Cantury Limiiad
31 L
V6
Fl
B 169
1G3AJ31N6KG330762
1969
CM
OU£MOOL£
CuU4«« Ciara
33 L
V6
Fl
B 169
JNlHJ0»P9CT37«777
1099
NISSAN
NISSAN
Macima
29 L
Vft
Fl
B 102
1G1LV13T6MY139222
1991
0J
CHEVROLET
B»r«lla
31 L
V6
Fl
0 194
162NE54U6MCS27I77
1 691
m
POMTIAC
GtmJAm LE
25 L
L4
F4
B 1®7
1GUF1IWXK71 74473
1 969
cu
acvnofr
CavaUr 224
29 L
V6
Fl
B 191
1G4CW&4C7X1639)19
1969
GM
BUCK
El*e«« Parti Awanua
38 L
V«
Fl
B 199
1G2WJ54T1MF245I32
1991
OA
POffllAC
Grand Pria
31 L
Vft
Fl
B 201
1G3H Y14C2KW366262
1909
GM
omsMoeu
Oatta 66 Royal* Brougham
39 L
Vft
Fl
B 211
1UEPM60T2MM631032
1991
FCPD
UfJEURf
Covgv 2 Dr. S*dan L3
5 0 L
Vft
EFI
B 212
183XG24K7KG107124
1909
CHRYSLER
DCDGE
Da^tona (U S |
2 5 L
L4
1 bbi
B 213
1G3AJ51N2KG32S301
1969
GM
CLDSMOOLE
Cutlaaa Ciara
33 L
Vft
F|
B 214
1P4GH64 R0MX579769
1991
CKRfStER
PLYMOUTH
Pfymovtfv Voya gar
33 L
Vft
ER
B 211
3G4AH$4NXMS903113
1991
CM
BUCK
Cantury Cwalom
33 L
Vft
Fl
8 218
f G2JC14K9M7S65174
1991
OA
POMTIAC
Sunbird
26 L
L4
Fl
B 219
2G4WD54LXU19&5715
1991
CM
BUCK
Ra^al
3 6 L
Vft
Fl
8 220
1B3XA4630MF564966
1991
a«ffsi£n
HUGE
Spirit
3 0 L
Vft
TBI
B 221
1G4HRS4C6MH401146
1991
ai
BUCK
LaSabr* Limilod
36 I
Vft
Fl
B 2 22
1GUC1HXKJ2I5IH
1969
OA .
CHEVROLET
Cavaliar
201
L4
Fl
B 223
1G4 AHS1N7KG4O06OO
1909
OA
BLHCK
Canfury Cuafem
33 L
Vft
Fl
B 224
lG2F$23T0ML2124«f
1991
OA
PONTtAC
F it •bird
31 L
Vft
Fl
B 226
B 227
B 230
2G1WLS4T1M9113477
1991
OA
OEV«X£T
Lumina
31 L
V6
Fl
IG3AJ5IW7KG320122
1999
GM
CL06M0QLE
CudtM Ciara
2 6 L
V6
Fl
0 232
1G3NFS4N4KU270366
1909
CM
OLDSMCWLE
Culaaa Calaia S
33 L
V6
Fl
6 236
1G4NV54N1MU2S9967
1991
CM
BUCK
Skylark
33 L
Vft
Fl
B 246
1FACPS0U4LA153137
1990
FCPD
FORD
Taurua L 4 Dr. Wagon
SOL
Vft
EFI
B 249
1FABP54Y4KAI90602
1969
FCFD
FCPD
Turn SHO
36 L
Vft
eh
B 254
1MEPM6O4OKHM4702
1909
FCPD
UEPtXJRf
Cougar 2 Or S«dan IS
36 L
Vft
EFI
Emiiaion
Country
Commtnt*
Control
Syai »m$
EVP/TWC/EGfVCa
CANADA
CAR 2 Ooot1 'opa Sadan
EVP/TWC/EGFVCU
USA
CAR 4 Dr $*4an
evp/twca:il
USA
CAR 2 Door Coupa Sad an
EVP/TWC/EGRCU
USA
CAR 4 Dr. Swlan
EVP/TWC/EGfVCii
USA
CAR
EVP/TWC/EGFVCa
CANADA
CAR 4 Doa S#dn
AlFUEVP/TWC/EGftCa
CANADA
MPV Wagar RAT
04001 -05000 GVWR t H
AIFVEVP/TWC/EGIVCU
CANADA
CAR
EVP/TWCyEGR/CLL
USA
CAR
EVP/TWC/EGR/CLL
JAPAN
CAR4 0) Stdan
E VP/TWC&GWCLL
USA
CAR
EVP/TWC/EGFVCU
CANADA
CAR 2 Door Coup* Sad an
AlfVEVfVTWC/EGR/CU
USA
CAR Zdr S«datn
EVP/TWC/EGftCLL
USA
CAR
AIR/EVP/TWC/EGR/CLL
USA
CAR
EVP/IWC/EGWCU
USA
CAR 4 Door Station Wagon
EVP/IWC«G»CLL
USA
CAR 4 Door S*zf«r
E VP/TWC/EGft'CLL
USA
CAR a Door Sadvt
EVP/TWC/EGFVCU
USA
CAR 2 Door Coup* S*dan
EVP/TWC/EGPVCU
USA
CAR 4 Door Sadan
EVP/TWC/EGR/CLL
USA
CAR 4 Door S«dan
AIR«VP/TWC/EGR/CL1
USA
CAR 4dr S+dan
EVP/TWC/tGFVCLL
USA
CAR 4 Door S*d«n
E VP/TWC/EGRrCLL
USA
CAR 4 DoarSadm
EVP/TWC/CLL
USA
CAR 4 Door Sedan
EVP/TWC/EGFfc-CLL
USA
CAR 4 Doe» S*d«r>
EVP/TWC/CLL
USA
CAR 4 Ooor S*dan
EVP/TWC/EGRfCU.
USA
CAR 2 Dm< C«vp* Sadan
AIR/EVP/TWC/EGR/CLL
USA
CAR
EVPfflrtC/EGFVClL
USA
CAR
EVP/TWC/EGWCa
USA
CAR 4 Door S*dan
EVPAWC/EGftCa
USA
CAR 4 Doa S*d«\
EVP/WC/CLL
USA
CAR 4 Ooct Sad at
EVP/TWC/CLL
USA
CAR 4 Doo Sad an
AlRJEVPft WC/EGFKC LL
JAPAN
CAR 4dt Sad an
EVP/TWC/EGFVCa
USA
CAR 2 Ddoi Coup*S*dan
EVPOWCrEGR/Ca
USA
CAR 4 Ooc» S*d«n
EVP/IWC/EGFVCLL
USA
CAR 2 Door Coupa S*dan
EVP/TWC«GfVCLL
USA
CAR 4 Dooi Sadan
EVP/TWC/EGFVCLL
USA
CAR 4 Ooa Sadan
EVP/TVKC/EGfVCLL
USA
CAR 2 Ooor Coupa S*dan
AIHEVP.TWC/EGR/CLL
USA
CAR
EVP/TWCVEGR/CLL
USA
CAR 2 Or Hafchba*
EVP/TWC/CLL
USA
CAR 4 Ooor Sadan
EVP/TWC/EGR/CU
USA
MPV Eitao4ad Wagon RAT
0500I-060CQ GVWR ft H
EVP/TWC/CLL
luEAOO
CAR 4 Ooor Sadan
EVP/TWC/EGR/CLL
USA
CAR2 Door Coup* S*dan
EVPfTWC/EGFVCLL
CANADA
CAR 4 Door Sadan
EVP/TWC/EGFVCLL
USA
CAR 4 Ct Sadan
EVP/TWC€GR'Ca
USA
CAR 4 fcx* Sadan
EVP/TWC/EGR/CLL
USA
CAR 2 Door Coup* S*dan
EVP/TWC/CLL
USA
CAR 4 Ooor Sadan
EVP/TWC/EG»CLL
USA
CAR 2 Door HatdibtcMjfiback
EVPiTWC€GFVCLL
CANADA
CAR 4 Doer S#dan
EVP/TWC/EGWCIX
USA
CAR 4 Ooor S*dar
EVP/TWC/CLL
USA
CAR 4 Dooi Sadan
EVP/TWC/CLL
USA
CAR 4 Doa Sadan
EVP/TWC€GR'CLL
USA
CAA
Evp/rwc/EG»ca
USA
CAP
EVP/TWCjEGR-CLL
USA
CAR
-------
Table 4-7 (Continued)
Sample
VIN
VIN
Model
Manufacturer
Mdte
Model
Engine
Fuel
Emission
Country
Commanta
Numb*
Confirmed
Ym
Induction
Control
B 257
by Decoder
System*
B 251
JH1HJ01P 4LT430661
1900
NBSAN
NISSAN
Muurie
29 L
V6
Fl
AIR/EVP/TWC/EGR/CLL
JAPAN
CAR 4dr Sedan
B260
UNUMB4SKV67J66S
1969
KH)
UTCOLN
Continetal GtvarwAy 4 Or Sedan
36 L
V6
EF1
EVP>TWC/EGfVCU
USA
CAR
B 261
2G4W054UM1044150
OJ
BUCK
Rsgd Limited
39 1
V6
Ft
EVP/TWCjEGR/CLL
CANADA
CAR 4 Door Sedvt
B 270
IG4HR54C3MH445M0
ot
BUCK
LeSebre Limited
36 L
V6
Fl
EVP/TWC/EGFVCLL
USA
CAR 4 Doo Sedvi
B 280
1FAPP0O45KH163635
1969
HJPO
R3BD
Ttandertwd Std 2 Dt Sedan
30 L
V6
EF1
EVP/TWC/tGR/CLL
USA
CAR
a 20*
JNlHJQtP$K1240Q?$
NISSAN
NSSAN
Maxima
29 1
V6
Fl
A(R/£YP/TWC/£GR/CLL
JAPAN
CAR 4dr Sedan
B 200
104HPS4C1MW424M7
Ol
BUCK
LiStU* Custom
3 9 L
V6
Fl
EVPrrWC/EGR/CLL
USA
CAR 4 Door Sedan
0 294
1MEPM604IKHC04264
RSC
VBCJH
Cougar 2 Dr Sedan L9
30 L
V6
EF1
EVP/TWC/tGR/Ca
USA
CAR
a 20s
1G3AJ51N0KG34O22O
(M
CUBMOBftE
Cudase Ciera
3 3 L
V6
Ft
EVP/TWC/CLL
USA
CAR 4 Ota Sedm
8 291
1LNBM01FOKY64O37O
FOFD
UNCOLH
town Car 4 Dr. Sedan
59 L
V6
2 btof
AlfVEVP/TWC/EGR/CLL
USA
CAR
B 200
1FAPP36X6KK1S3561
BOF©
FCRD
Tempo 4 Dr. Sedan GL
23 L
L4
CFI
EVP/TWC/EGR«a
USA
CAR
6 307
tGUV14W6KY160794
Ol
ocvnatr
Bareiia
20 L
V6
Fl
EVP/TWC/EGR/CU
USA
CAR 2 Door Coupe Sedan
1G3AJ5ltt7KG3324l0
Wgn GS
30 L
V6
ER
EVP/TWC/EGR/CU.
USA
CAR
B 323
1FABP55UIKA106S99
1969
RSC
KPD
Taurus L 4 Di Wagon
3 0 L
V6
er
EVP/TWC/EGFVCLL
USA
CAR
B 32#
1C3X Y56R3LD910004
1990
CHRTStER
ovrisim
imperial
2 2 L
LA TURBO
Fl
AIR/CVP/TWC/EGR'CLL
USA
CAR 4 Or Sedan
B 327
1P3XA563CMF541006
CHRVSIER
R.YMCUTH
Acclaim LE
3 0 I
V6
TBI
EVP/TWC/EGR/CU.
USA
CAR 4 Dr. Sedan
B 321
2MEPM36X5KB6065S*
FCFO
l/BCURf
Topaz 4 Ol. Sedvt GS4 WDR
2 3 L
L4
CFI
E VP/TWC /EGfVC U.
CANADA
CAR
B 330
IFABP52U4KA140006
KH>
FOFC
Te«yv» GL 4 Or. Wagon
3 0 L
V6
ER
EVP/TWC/EGR/CU.
USA
CAR
B 331
1G3HN54C2KW334I47
OJ
OLD6MOOU
Oelie 66 Roy ale
3 6 L
V6
Ft
EVP/TWC/EGRTCLL
USA
CAR 4 Door Sedan
B 332
2G3AJ54N6M2333116
CM
CtDSMOBU
Cudase Ciere
3 3 L
V6
Ft
EVP/TWC/CLL
CANADA
CAR 4 Doct Sedan
B 333
1FABP50O1KG155I27
(CR)
RVC
Teurue 14 0« Wagon
2 5 L
L4
CFI
AIR/€VP/f WC/EGFVCLL
USA
CAR
B 334
2G4WBMT2M141S297
1991
EGRrCU-
CANADA
CAR 4 Door S*d«n
B 330
1G4AH54N4M641S191
CM
BUCK
Century Custom
33 L
V6
Fl
EVP/TWC/CLL
USA
CAR 4 Door Sedan
B 330
1YVGO22B0MS1OS315
MUDA
MAZDA
626
2 2 L
L4
Fl
EVP/TWC/EGWCa
USA
CAR 4 Or. Sedan
B 340
1FABP5346KA149225
FOFD
FOFD
Taurus LX 4 Or Wagon
3 0 L
V6
En
EVP/TWC/EGRrCa
USA
CAR
B 341
1B3XP26O2MN554504
DCOGE
Shadow
23 L
i.4
Fl
EVP/TWC/EGWCa
USA
CAR 4 Dr Hatefcbe*
8 342
1YVGD22B6M512&617
MAZDA
MAZDA
626
22 1.
L4
Fl
CVP/TWC/EGFVCLL
USA
CAR 4 Dr. Svdan
B 357
2B4FK25K6KR225696
ckmsler
OdXE
Dodge Car aw an
25 L
L4
ER
AIR/6 VP/TWC/EGR/CLL
CANADA
UPV Wagon FIAT 04001 -05000 GVWR 6 H
B 372
1C3BC6637KD543222
awskSR
cwwler
New Yorfcer Landau
3 0 L
V6
TBI
EVP/TWC/EGR/CLL
USA
CAR 4 Or Sedan
0 374
1G4AH51N6K1421352
at
BUCK
CoAlury CuiImi
3 3 L
V6
Fl
EVP/TWC/CLL
USA
CAR 4 Door Sedan
a 3»i
1G2HXS4C XKW263004
GM
P0NT1AC
BonneviHe LE
3 6 L
V6
Fl
EVP/TWC/EGttCLl
USA
CAR 4 Doer Sedan
a 303
1G4HRS4C1MH474455
1901
CM
BUCK
LeSebre Umitad
3 8 L
V6
F>
EVP/TWC€GFVCLi
USA
CAR 4 Doer Sedan
B 304
1G3HN54C2KW324623
Ol
OU36MCOU
DeJu M Ro
3 6 L
V6
Fl
EVP/TWC«GFVCU
USA
CAR 4 Door Sedwt
B 300
2GIWL54T3L6124656
1900
Of
OCVACXET
Luffiina
3 1 L
V6
Fl
EVP/TWC/EGRrCa
CANADA
CAR 4 Door Setfwi
B39T
1G t JF t1 WOK?173624
Qi
OCVROLET
CavaKet Z24
2 6 L
V6
Fl
EVPiTWCiEGIVCU.
USA
CAR 2 Door Coup* Sedan
B 300
1N4GB22S1KC766337
MSSAN
NSSAN
Sentoe
16 L
L4
Fl
AIR«VP/TWC/EGR/CU
USA
CAR 26> Sedan
B 400
1G3HN54COKH300390
1900
CM
0LD6MCBLE
Delta 66 Roy4e
3 6 L
V6
Fl
EVP/TWC/EGR'CLL
USA
CAR 4 Door Sedan
B 401
1G4HPMC1KH536052
I960
CM
BUCK
LeSebre Cuttom
36 L
V6
Fl
EVP/TWC/EGWCa
USA
CAR 4 Door Sedan
B 403
1G3AJ$11W0KG317014
-------
4.4
Factors Affecting the Data
In addition to the vehicle and driver factors already described that could
affect the driving pattern data, weather conditions could also affect the way people drive.
As part of this study, an effort was made to document pavement conditions during the
days vehicles were driving with instrumentation on-board.
Table 4-8 summarizes the estimated driving conditions in Spokane and
Baltimore. The table gives the solicitor's estimate of the pavement condition for each
day of instrumentation. Because there were some days where the solicitor did not
document pavement conditions, the National Climatic Data Center (NCDC)
meteorological data collected at the Spokane Airport and downtown Baltimore are also
shown. The Spokane Airport is on a plateau above the city of Spokane, which is in a
valley; therefore, the meteorological data at the airport and the actual road conditions in
the city do not necessarily agree. In addition, it is likely that drivers who had
instrumented vehicles did not drive only in the city of Spokane, but may have driven to
surrounding areas, including to Mt. Spokane, where snow was present during the study.
Data from the downtown Baltimore station is probably representative of
conditions at the Baltimore City station, which is two miles north of downtown, but not
necessarily of the Baltimore County station, which is 10 miles north-northeast of
downtown Baltimore. In addition, it is likely that drivers who had instrumented vehicles
did not drive only in the city of Baltimore, but may have driven to surrounding areas —
most notably to Washington, D.C.
Table 4-8 shows NCDC meteorological data for precipitation, snowfall, and
snow depth in inches. The overall estimates of pavement condition given in Table 4-8
are based on the solicitor's observations and on the NCDC data.
4-54
-------
Table 4-8
Road Condition Estimates
City
Date
Solicitor Observations
NCDC Meteorological Data
Overall
Estimate
of Pavement
Condition
Pavement
Condition
Comments
Precipitation
(in.)
Snowfall
(in.)
Snow Depth
(in.)
Spokane
03Feb92
Dry
Sunny,35
0
0
0
Dry
Spokane
04Feb92
Dry
Sunny, 40
0
0
0
Dry
Spokane
05Feb92
Dry
Foggy, 30
<0.005
<0.05
0
Dry
Spokane
06Feb92
Dry
Sunny, 30
0
0
0
Dry
Spokane
07Feb92
Dry
Sunny, 35
0.08
0
0
Dry
Spokane
08Feb92
Wet
Rainy, 35
0.17
1.4
0
Wet
Spokane
09Feb92
<0.005
0
<0.5
Dry
Spokane
10Feb92
Dry
Sunny,30
0
0
0
Dry
Spokane
1lFeb92
Dry
Sunny,38
<0.005
0
0
Dry
Spokane
12Feb92
Wet
Cloudy, 35, streets damp
<0.005
0
0
Dry
Spokane
13Feb92
Wet
Cloudy, 38, streets damp
0.03
0
0
Dry
Spokane
14Feb92
Wet
Rainy, 38
0
0
0
Dry
Spokane
15Feb92
Dry
Sunny,40
0.02
0
0
Dry
Spokane
16Feb92
0
0
0
Dry
Spokane
17Feb92
.
0.09
<0.05
0
Dry
Spokane
18Feb92
Wet
Rainy, 34
0.66
<0.05
0
Wet
Spokane
19Feb92
Wet
Sunny, 36, streets damp
0.01
<0.05
0
Dry
Spokane
20Feb92
Wet
Rainy, 34
0.44
<0.05
0
Wet
Spokane
21Feb92
Wet
Cloudy, 38, streets damp
0.06
0
0
Dry
Spokane
22Feb92
<0.005
0
0
Dry
Spokane
23Feb92
0.11
0
0
Dry
Spokane
24Feb92
0.09
0
0
Dry
Spokane
25Feb92
<0.005
0
0
Dry
Spokane
26Feb92
0
0
0
Dry
Spokane
27Feb92
0
0
0
Dry
Spokane
28Feb92
<0.005
0
0
Dry
Spokane
29Feb92
<0.005
0
0
Dry
Spokane
01Mar92
0.10
0
0
Dry
Spokane
02Mar92
0.005
0
0
Dry
-------
Table 4-8
(Continued)
City
Solicitor Observations
NCDC Meteorological Data
Overall
Estimate
of Pavement
Condition
Date
Pavement
Condition
Comments
Precipitation
On.)
Snowfall
(in.)
Snow Depth
(in.)
Baltimore
05Mar92
Dry
Cloudy
0
0
0
Dry
Baltimore
06Mar92
Dry
Morning fog, drizzly afternoon
0.07
0
0
Dry
Baltimore
07Mar92
Wet
Rainy
0.85
0
0
Wet
Baltimore
08Mar92
0
0
0
Dry
Baltimore
09Mar92
Dry
Sunny
0
0
0
Dry
Baltimore
10Mar92
Dry
Sunny
0.09
0
0
Dry
Baltimore
1 lMar92
Dry
Flurries, high winds
0
2.8
0
Dry
Baltimore
12Mar92
Dry
Sunny, cold
0
0
0
Dry
Baltimore
13Mar92
Dry
Sunny, windy
0
0
0
Dry
Baltimore
14Mar92
Dry
Sunny
0
0
0
Dry
Baltimore
15Mar92
0
0
0
Dry
Baltimore
16Mar92
Dry
Sunny
0
0
0
Dry
Baltimore
17Mar92
Icy/Wet
Morning sleet, afternoon rain
0
0
0
Dry
Baltimore
18Mar92
Dry
Sunny
0.45
0
0
Wet
Baltimore
19Mar92
Icy/Wet
Rain/sleet
0.2
0
0
Wet
Baltimore
20Mar92
Dry
Sunny
0
0
0
Dry
Baltimore
21Mar92
Dry
Sunny
0
0
0
Dry
Baltimore
22Mar92
0.2
0
0
Wet
Baltimore
23Mar92
Dry
Sunny
0
0
0
Dry
Baltimore
24Mar92
Dry
Sunny
0
0
0
Dry
Baltimore
25Mar92
Dry
Sunny
0.03
0
0
Dry
Baltimore
26Mar92
Wet
Rainy
1.89
0
0
Wet
Baltimore
27Mar92
0.01
0
0
Dry
Baltimore
28Mar92
0
0
0
Dry
Baltimore
29Mar92
0
0
0
Dry
Baltimore
30Mar92
0
0
0
Dry
Baltimore
31Mar92
0
0
0
Dry
Baltimore
01 Apr92
0.06
0
0
Dry
Baltimore
02Apr92
0
0
0
Dry
-------
5.0
DATA QUALITY CHECKING AND ARCHIVING
The next step after collecting field data is checking the data quality and
archiving it in a suitable format in preparation for data analysis. These activities include
writing computer programs to highlight suspect data points for manual inspection,
manual inspection of the data, editing the data, and writing computer programs to
convert the raw and edited data into SAS data sets for analysis.
In this section, the data quality checking methodology is outlined and
sample plots are used to describe the method. The rules developed during the data
quality check are summarized and the results of the data quality check are presented.
5.1 Data Quality Checking Methodology
Checking of data quality was intended to be accomplished as soon as
possible after the field data arrived for each vehicle. Because of the rate at which the
data arrived in Austin and the large amount of work required to check it for each
vehicle, data from the field for almost all the vehicles were examined many days after
arrival. Nevertheless, validation and inspection of the data for the first few cars was
used to detect systematic problems in the installation and removal of dataloggers and
transducers in the field.
In general, the data from the field were checked using a multiple step
methodology that combined highlighting suspect data points using SAS software
programs with manual editing and examination of graphical output by engineering
personnel. Figure 5-1 shows a flow diagram of the steps used to check the quality of the
raw data and to load it into the final edited SAS data files. The following discussion
presents the details of the activities outlined in Figure 5-1.
5-1
-------
Figure 5-1.
Data Quality Checking Methodology
5-2
-------
The raw data for each vehicle were received in Austin from the field on
3V2-inch floppy disks. The data had been compressed in the field using ARJ software.
With the floppy disk was a photocopy of the data packet used in the field.
The first step in checking the data quality was to inspect the data packet.
This inspection indicated whether the installation and removal of datalogging equipment
on the vehicle had been successful and whether the data had been successfully
downloaded from the datalogger. Any difficulties with the instrumentation of the vehicle
were noted in the data packet.
The compressed data from the floppy disk was uncompressed, again using
ARJ software and an ASCII file was created. The name of the file was made up of the
last eight digits of the vehicle identification number. For the purposes of this discussion,
VIN.RAW will indicate raw data files. This file was placed on the mainframe computer
for data quality checking and conversion to SAS format. The raw data file was examined
using an editor to determine if any obvious errors could be found. In a few instances,
the beginning of the data file was corrected where spurious keystrokes were mistakenly
inserted by data downloading personnel. The second-by-second numerical data were
briefly reviewed and, in all but a few cases, were found to be clean.
The next major step in the data quality checking methodology was to run
each of the VIN.RAW files through a SAS program called READ.SAS. READ.SAS
was written to make several diagnostic checks of the raw data, to present the results of
the driving patterns to engineering personnel in a way that would highlight suspect
values, and to create files that could be used to correct the raw data files when personnel
decided that individual logged values needed to be corrected. READ.SAS makes two
files, a listing, and several diagnostic plots. The two files are READ.ERR and
READ.WOT.
5-3
-------
READ.ERR contained a listing of all of the values that READ.SAS
considered suspect, based on criteria in the READ.SAS program. The logged values five
seconds before and five seconds after suspect data points were also included in
READ.ERR so that the editing personnel could use surrounding data information to
judge the suspect data value. The suspect data values and their neighboring values are
called a group. The criteria that READ.SAS used to flag suspect data values are given
in Tables 5-1 and 5-2 for 3-parameter and 6-parameter data, respectively. The
READ.ERR files contained the logged data values in duplicate columns. This permitted
the second of the columns to be edited to the correct values, so that when the
READ.ERR file was merged with the raw data file, the corrections would be applied to
the raw data file without disturbing the original values in the raw data file. In this way,
the corrected data could be analyzed; however, if an investigator wants to examine the
edits made to the raw data to get the corrected data, he can examine READ.ERR,
because it contains both the original suspect values and the values they were changed to.
READ.SAS also made a file called READ.WOT. These files were made
for the 3-parameter data only, and the files contained all observations made when the
datalogger was in the calibration mode. That is, whenever the installation and removal
crew were driving the vehicle, the data logged under those circumstances were put in the
READ.WOT file. During the calibration drive during installation and during the
removal drive, wide-open throttles were performed. The READ.WOT files were later
edited so they contained only the vehicle data obtained during wide-open throttle
operation. These files would be useful for evaluating how the vehicle driver drove the
vehicle, compared with a measure of the ultimate performance of the vehicle.
READ.SAS also created a listing file that contained a number of diagnostic
aids. This included a listing of any parity errors that may have occurred in the data as
logged on the datalogger, a comparison of the difference in time between the time stamp
at the beginning of a trip and the time stamp at the end of a trip with the number of
seconds of observations between the time stamps, a listing of the start and end of each
5-4
-------
Table 5-1
Suspect Observation Flagging Parameters for
3-Parameter Data
| Accel | > 4 m/s2
| A RPM| > 1400 rpm/s
(A MAP[ > 60 kPa/s
Speed > 40 m/s
RPM > 6000 rpm
RPM < 300 rpm
MAP < 10 kPa
MAP > 100 kPa
The first or last speed of a trip > 0 m/s.
Speed = 0 m/s and RPM = 0 rpm
MAP > 80 kPa and Speed = 0 m/s
5-5
-------
Table 5-2
Suspect Observation Flagging Parameters for
6-Parameter Data
| Accel | > 8.95 mph/s (4 m/s2)
| A RPM | > 1400 rpm/s
Speed > 89.5 mph (40 m/s)
RPM > 6000 rpm
RPM < 300 rpm
For MAP systems: MAP < 10 kPa
MAP > 100 kPa
| A MAP| > 60 kPa
For MAF systems: MAF > 500
For LV8 systems: LV8 > 255
Throttle position > 103%
Throttle position < -3%
Coolant Temperature > 150°C
Coolant Temperature < -25°C
| A Coolant Temperature | > 20°C/s
4> > 1.5
15%
The last speed of a trip > 1 mph.
5-6
-------
trip for all the trips, summary statistics on speed, RPM, manifold absolute pressure,
acceleration, change in RPM, and change in manifold absolute pressure, a listing of all
the suspect data values found in the file, and a listing of all the groups associated with
each suspect data value. This group listing was similar to the READ.ERR file, but
included additional diagnostic aids that identified the start and end of each trip and the
suspect observation.
Finally, READ.SAS made diagnostic plots for the data file. A copy of the
diagnostic plots for READ.SAS for one vehicle appear in Appendix A. This series of
plots is for a vehicle with a 3-parameter datalogger. Figure A-l is a plot of the number
of groups, identified as having suspect data values, versus the date and time. The
uncalibrated speed in meters per second for the vehicle during the installation calibration
drive is given in Figure A-2. Note that the speed during the first half of the drive pins at
40.95 m/s because the speed transducer had not yet been calibrated. The second half of
the calibration drive contains a wide-open throttle acceleration. Figure A-3 shows the
removal drive, during which there is an additional wide-open throttle acceleration.
Wide-open throttle accelerations can be determined by consulting the time recorded for
the wide-open throttle acceleration in the data packet and by considering the manifold
absolute pressure logged in the data file.
The number and time of occurrence of the trips are given in Figures A-4
and A-5, versus date and time and cumulative engine-on time, respectively. The date
and time of trips during the entire instrumentation period of approximately a week can
be seen in Figure A-6. Spaces between the vertical peaks usually indicate periods when
the engine was not operating. Figure A-l shows vehicle speed versus cumulative engine-
on time. This plot does not show any engine-off time; therefore, horizontal lines at zero
speed indicate idles. Engine RPM versus cumulative engine-on time is given in Figure
A-8. Data values at engine starts and stops sometimes went to zero RPM. These can be
seen in the figure as spikes pointing towards zero. In the data editing process, these
values were removed from the beginning and end of trips.
5-7
-------
The trend of manifold absolute pressure with cumulative engine-on time is
shown in Figure A-9. Values around 20 kPa represent a throttle that is mostly closed,
and values near 100 kPa represent operation with the throttle mostly open. Values that
spike towards 0 kPa can occur at the beginning and end of trips. These are routinely
edited out in the data quality checking process. Vehicle acceleration was computed by
subtracting adjacent speed values and plotting in Figure A-10 versus cumulative engine-
on time. An acceleration of zero represents driving at a constant speed or idling,
whereas positive values mean an acceleration and negative values mean a deceleration.
The rate of change of engine RPM is plotted in Figure A-ll. The rate of change of
manifold absolute pressure is plotted in Figure A-12. Figure A-13 shows the second
derivative of speed, which is the jerk.
Three additional diagnostic plots were used to examine vehicle and engine
operation between pairs of parameters. In Figure A-14, the acceleration plotted against
the speed shows the regime of vehicle operation on the road. Because of the
characteristics of transmissions, higher accelerations can be obtained at low speeds, when
the transmission is in a low gear, whereas at high speeds more moderate accelerations
are typical, because the transmission is in a high gear. This behavior produces a
characteristic triangular shape in the acceleration vs. speed plots, which was seen
throughout the data of all the vehicles. Figure A-15 shows a plot of engine RPM versus
manifold absolute pressure. The region of high density data between 20 and 100 kPa
and between 1000 and 3000 rpm was typical of most vehicles in the data sets.
Turbocharged vehicles frequently had manifold absolute pressures that exceeded
atmospheric pressure (100 kPa). These high pressures were correctly logged by the
dataloggers. In Figure A-16, the ratio of speed to RPM was plotted against vehicle
speed. The ratio of speed to RPM is proportional to the overall gear ratio of the vehicle
at different speeds. In the example shown in the figure, the vehicle had a 5-speed
manual transmission. This is characterized in the figure by data that lie on 5 horizontal
lines. Close examination of the data indicates the speeds at which shifts were made as
speed increased.
5-8
-------
The same approach was used to validate and identify suspect data values in
the 6-parameter data.
All of these diagnostic plots were used to determine if outliers in the data
existed. The same plots were made using later programs in the data quality checking
methodology and compared with these first plots to determine if the data edits improved
the visual appearance of the data as plotted.
Following these diagnostic plots, additional plots were made of the logged
data for each group of data values surrounding suspect data values. These plots enabled
the data editor to visualize the region around suspect data values and to suggest the best
way to handle suspect data values.
The outputs from READ.SAS were used by data file editors to consider
the suspect value observations flagged by READ.SAS. If the editor thought a change
was justified, he marked the second printed value in the listing READ.LST. The
guidelines used to help the editor make these decisions are reviewed in Section 5.2. The
data editor considered the diagnostic plots and the diagnostics given at the beginning of
the READ.LST listing. The basic concept followed was that logged values that were
reasonably possible for the vehicle to produce were left alone. If it was judged that the
logged values were not possible, they were either changed to estimates of the correct
values or changed to missing values.
If, in the judgment of the data quality checking team, the logged data
exhibited characteristics that would bias the overall results of the study and that were not
possible to check without fabricating a considerable amount of data, the entire file for
that vehicle was considered suspect. These files were set aside and not considered
further. Thus, all of the data presented and analyzed for this final report are the
cleanest and most reliable data collected in the study. In the future, the suspect files will
5-9
-------
be re-examined to determine if portions of the data can be recovered for inclusion in the
data set.
After the initial markup of the READ.LST listing, a second data editor
made a quality control check of the markups to the listing. If the second editor
disagreed with some of the markups, a conference on the data file was held. Otherwise,
the data file passed to the next step of editing.
In the next step, the marked up READ.LST listing was used by the file
editor to make edits on the computer to the READ.ERR file. In addition, the file editor
examined the READ.WOT files and the data packets and determined which portion of
that file represented the best example of wide-open throttle accelerations for the
calibration drive during installation and the calibration drive during removal of the
datalogging equipment. Since wide-open throttle accelerations were performed only on
vehicles equipped with 3-parameter dataloggers, this edit could only be performed for 3-
parameter data. The identity of the wide-open throttle accelerations in the data was
determined by the rate of increase of vehicle speed, an elevated manifold absolute
pressure, and the time recorded in the data packet by the drivers performing the wide-
open throttle acceleration. All other driving data obtained during the calibration drives
were deleted from the READ.WOT files.
In the next step of the data quality control checking process, another
program called CHECK.SAS was run. This program took the raw data in VIN.RAW
and the edits made to READ.ERR and READ.WOT and produced a corrected data file,
which includes all the edits. In addition, CHECK.SAS made all of the same diagnostic
checks that were included in READ.SAS and produced a listing called CHECK.LST and
all of the same diagnostic plots. The purpose of these diagnostic aids was to enable the
data editing team to ensure that the edits to the data entered in READ.ERR were made
correctly and were reasonable.
5-10
-------
Accordingly, the data editing team took the outputs of CHECK.SAS and
examined them by considering them by themselves and also by comparing them with the
outputs of READ.SAS to see if improvements were made. If the examination
determined that additional edits needed to be made, the READ.ERR file was re-edited
and CHECK.SAS was run again until the data editing team was satisfied.
In the final stage of the data quality checking process, a program called
LOAD.SAS created the final edited SAS data files in the required data archiving format,
which is described in Section 5.3. The data for all vehicles were concatenated with each
other and the units of the recorded values were made consistent for both 3-parameter
and 6-parameter data.
5.2 Data Correction Guidelines
The criteria used to decide if an individual data value was reasonable or
unreasonable were developed as the data quality checking procedures were taking place.
This was done because before any data were collected, the types of data value errors and
their characteristics were not known. The characteristics of the 3-parameter datalogger
and the 6-parameter datalogger were different. In addition, some modifications to the 3-
parameter datalogger software were made between the Spokane and Baltimore trips.
These modifications changed and improved the characteristics of the data collected in
Baltimore.
The sections that follow present, in some detail, the guidelines used to edit
the data for the Spokane 3-parameter, the Baltimore 3-parameter, and the 6-parameter
datalogging systems. It was found that many of the suspect data values occurred during
engine startup and engine shutdown; therefore, the criteria used by the dataloggers to
determine the beginning and end of data logging episodes is important to outline before
discussing the data editing criteria. The criteria were described in Section 2.1.1.
5-11
-------
5.2.1 Guidelines for Spokane 3-Parameter Data Editing
The following guidelines were used to edit the data obtained from the 3-
parameter dataloggers used in Spokane and from the first 11 3-parameter vehicles (S002,
S007, S010, S016, S017, S019, S037, S048, S050, S052, and S066) instrumented in
Baltimore.
Speed
1. For driving episodes with nearly constant speed, constant RPM, and
constant MAP, when a sudden drop in speed occurs for one second
followed by a resumption of the previous speed or a speed close to
it, the speed spike value was replaced by an interpolation of the
points adjacent to it.
2. A characteristic speed spike was observed in the Spokane 3-
parameter data and was linked to the datalogger software. This
error is characterized by a speed about 2.7 times the actual speed
for one observation and then by a speed about 0.61 times the actual
speed for the next observation. When the spike produced an
acceleration value in excess of 4 m/s2, these incorrect values were
replaced by corrected values calculated from the logged values
divided by their corresponding correction factor. This error
occurred once approximately every 20,000 seconds of driving.
3. For a non-zero speed value at the beginning or end of a trip
adjacent to other non-zero speeds, the speed value was left
unchanged.
4. For a non-zero speed value at the beginning or end of a trip
adjacent to two or more zero speed values, the speed value was
considered to be a startup or shutdown spike and was set to zero.
5. If the first speed value at the beginning of a trip was large (>5 m/s)
and was followed by non-zero speeds that were much lower, the first
speed value was considered to be a startup spike of unknown true
value and was changed to a missing value.
6. If the end of a trip had two or more non-zero speed values preceded
by zero speed values, all values were left alone.
5-12
-------
7. A datalogger shutdown that resulted in one or more seconds of lost
data could be identified by examining the vehicle and engine speeds
at the suspected end of trip and corresponding beginning of the next
trip. If the speed was non-zero and the RPM was idle rpm or
greater, then this was probably a datalogger shutdown error. If the
engine RPM dropped below idle (<500 rpm) and the MAP went to
atmospheric pressure, then the vehicle may have stalled, and this
could not be considered a datalogger error. But if the MAP was
still low, the datalogger was probably shut off because of a low
quality RPM signal.
RPM
1. If the RPM value was greater than 2500 rpm for the first
observation at the beginning of a trip and then the RPM value
dropped to normal idle (<. 1000 rpm), the RPM value was considered
to be a startup spike. It is unlikely that such high RPMs would be
experienced for an engine for only one second at startup. This error
was changed to a missing value because the real RPM was
unknown.
2. A characteristic RPM spike was observed in the Spokane 3-
parameter data and was linked to the datalogger software. This
error was characterized by an RPM about 2.7 times the actual RPM
for one observation and then by an RPM about 0.61 times the
actual RPM for the next observation. When the spike produced an
RPM charge greater than 1400 rpm/s, these incorrect values were
replaced by corrected values calculated from the logged values,
divided by their corresponding correction factor. This error
occurred once approximately every 20,000 seconds of driving.
MAP
1. If there was a very low MAP (<5 kPa) during the first second of a
trip followed by an MAP near atmospheric pressure, the first MAP
data value was considered a datalogger startup spike and was
changed to a missing value.
2. A MAP value was flagged if the absolute value of the change in
MAP was greater than 60 kPa/s. Such a MAP change could occur
5-13
-------
frequently during shifting while accelerating and general revving of
the engine to avoid a stall. The MAP change should be
accompanied by a similar pattern in RPM. If it was not, and the
MAP change occurred for one second only, then it was considered a
MAP spike. In this case, the MAP value was changed to a missing
value.
5.2.2 Guidelines for Baltimore 3-Parameter Editing
The following guidelines were used in addition to the guidelines used for
the Spokane data for all Baltimore 3-parameter data except for data from the first 11
Baltimore 3-parameter vehicles.
1. If the RPM was less than 300 rpm for a few seconds at the
beginning of a trip, these observations for all parameters were
deleted from the file.
2. If at the end of a trip, RPM equaled 0 rpm and speed equaled 0
m/s and MAP was rising, the entire observation was deleted for all
parameters. In this case, the vehicle had been shut off, but the
MAP had not reached atmospheric pressure yet.
Logic used for the 3-parameter datalogger in Baltimore produced a slightly
different kind of data for the beginning and end of trips than that produced by the 3-
parameter datalogger in Spokane and the 6-parameter datalogger. This can be
illustrated by considering a number of different shutdown and startup scenarios, all of
which were observed in the Baltimore data.
The most common beginning and end of a trip are shown in Figure 5-2a.
For a normal trip, the datalogger turns on as soon as the manifold absolute pressure
drops below 80 kPa. A short time after this occurs, the engine RPM increases from the
cranking RPM up to idle RPM. When the driver begins moving the vehicle, speed
increases above 0.1 m/s, and the vehicle is in motion. At the end of a trip, the driver
5-14
-------
a) A normal trip b) Coasting with engine off
Lri
¦
t—''
<-J\
80 kPa
80 kPa
300 rpm
300 rpm
Trip Continues Trip Continues
c) Engine dies and is restarted with no stop of motion d) Engine dies and is restarted, but motion stops only after a restart
Figure 5-2. Vehicle/Engine Operating Scenarios Relevant to Datalogger Operation
-------
stops the vehicle and turns off the engine. During this period, vehicle speeds drops
below 0.1 m/s, the RPM drops abruptly below 300 rpm; however, the MAP may still be
below 80 kPa. Thus, the end of the datalogging period occurs when the MAP exceeds 80
kPa. Thus, it is clear under a normal circumstance that it is actually the MAP that is
turning the datalogger on and off. One advantage of this logic is that it allows the data
to be collected during cranking of the engine. Subsequent examination of the data can
determine whether the engine was started on the first cranking attempt or whether it
required several cranking attempts. If the MAP is connected improperly to the
datalogger or the MAP sensor does not respond, trip beginning and end will be
determined by the RPM signal exceeding 300 rpm. Finally, if the RPM signal is
unreliable or not working, the actual motion of the vehicle can start the datalogger, as
provided for by the logic in the datalogger program.
Figure 5-2b shows another type of driving behavior that was observed in
the Baltimore data on several occasions. Some drivers either deliberately or accidentally
experienced an engine shutdown while the vehicle was still in motion. Some time later,
the engine was restarted and another trip began. Figure 5-2b shows that for the
Baltimore 3-parameter datalogger, the end of the datalogging episode would occur when
the speed of the vehicle became less than 0.1 m/s. The datalogger would start again
when the cranking of the engine caused the MAP to drop below 80 kPa. For the
6-parameter datalogger and for the 3-parameter datalogger in Spokane, the end of the
trip would be somewhat different, since it was controlled by the engine RPM dropping
below 300. A shutdown of the datalogger at this point would result in a non-zero speed
at the end of the trip. An editor examining such data would not be able to determine
whether the vehicle had coasted after the datalogger shut down or whether the data
values for the speed were not accurate at all. With the 3-parameter datalogger in
Baltimore, the speed of the vehicle could be followed as it coasted down with the engine
off to the end of the trip.
5-16
-------
Another type of event that happened on a few occasions in Baltimore is
shown in Figure 5-2c. In this instance, the engine died while the vehicle was in motion.
However, the engine was restarted before the vehicle came to a complete stop. Because
of this, the 3-parameter datalogger in Baltimore did not turn off at all. In the data
checking process, this was viewed as a continuation of an existing trip. For the
6-parameter datalogger and the 3-parameter datalogger in Spokane, this same situation
would have been recorded as the end of one trip and the beginning of another trip.
However, both the speeds at the end of the trip and the beginning of the next trip would
be non-zero. This would have been viewed with concern by the data checking team,
since what happened between those two events was not logged and could not be known.
A more unusual scenario is shown in Figure 5-2d. In this instance, the
engine died and the vehicle began slowing down, but the driver of the vehicle was able
to restart the engine before the vehicle came to a stop. But the vehicle did come to a
stop after the engine was restarted. Again in this instance, the Baltimore 3-parameter
datalogger did not shut down during this event because at all times at least one of the
three sensors was in the running state. If the 3-parameter Spokane datalogger or the 6-
parameter datalogger had experienced this situation, it would have recorded the end of
the one trip and the beginning of another trip when engine RPM passed through 300
rpm.
These examples show that the beginning and the end of a trip is not always
a clear-cut event. But whatever the event may be, the logic used for the Baltimore
3-parameter datalogger was able to record the information throughout such events so
that the investigator could determine what was happening to the automobile.
5.2.3 Guidelines for 6-Parameter Data Editing
The 6-parameter datalogger did not experience problems at the beginning
of trips because of datalogger startup transients, because the 6-parameter datalogger did
5-17
-------
not record any values for all six parameters during the first five seconds after engine
startup. In addition, the equivalence ratio determined by the exhaust system oxygen
sensor did not record any data on the datalogger for the first 60 seconds after engine
startup. All of these data points are represented as missing values in the data files.
The following guidelines were used to edit the 6-parameter data:
1. If any one of the six parameters displayed a spike at the end of a
trip, the spike value was replaced with a missing value.
2. If the end of trip speed was non-zero and it was preceded by larger
non-zero speed values, the speed value was left alone.
3. If the end of trip speed value was non-zero and it was preceded by
two or more zero speeds, the value was changed to zero.
4. If more than one of the six parameters displayed a spike at the end
of a trip, the entire last observation of the trip was deleted. This
type of spike value set was observed frequently, and it was
presumably the result of datalogger shutdown transients at the end
of trips.
5. If the value of the equivalence ratio for the last second of a trip had
a spike value near 2.3, the entire last second of the trip was deleted,
since this value for equivalence ratio indicated that the datalogger
was in the process of shutting down and values for other parameters
would not be reliable.
6. If RPM, MAP, and throttle position were all relatively constant and
the equivalence ratio changed significantly (A<{> >0.3/s), the value of
the equivalence ratio was replaced by an interpolation of adjacent
values.
7. Spikes in equivalence ratio occurring during transients of the other
five parameters were left alone, because it was not known what the
response of equivalence ratio would be during transient behavior of
the vehicle.
5-18
-------
5.3
Data Archiving Format
Running the SAS program LOAD.SAS produced four SAS data set files
that contain all of the driving pattern data. The four files are the vehicle file, the
operation file, the wide-open throttle file, and the error file. These four files were
analyzed to produce the desired results of this study.
The data archiving formats of these four files are given in Table 5-3. The
files have the same formats for both 3-parameter and 6-parameter data. For data that
do not exist because of the type of datalogger on which they were obtained, the values
for observations will have missing values designated by a period in accordance with SAS
protocol. For example, data observations collected on the 3-parameter datalogger will
always have missing values for throttle closed position voltage, throttle open position
voltage, MAF, LV8, throttle position, coolant temperature, and equivalence ratio.
The vehicle file (VEH.SSD) contains descriptions of the test vehicles that
were instrumented. The file contains one observation for each vehicle. Thus, the
vehicle file has 216 observations. The VIN in all four files is the parameter used to
connect the four files for data analysis. The VIN in these files is the VIN as it was read
from the vehicle in the field. Many of the field VINs have transcription errors. The
correct VINs were determined after returning from the field, but the VINs in the data
files were left uncorrected to avoid database problems. The corrected VINs are given
elsewhere in this document. The model year, make, model, transmission type, and
odometer reading have values determined by the on-site personnel when they inspected
the vehicle. Load measurement type has a value of MAP, MAF, or LV8, depending on
which type of load was used by the datalogging system. MAP was the load measurement
type for all 3-parameter logged data. The 6-parameter datalogging system used any of
the three load measurement types. The throttle closed position voltage and the throttle
open position voltage were recorded in the data packets for 6-parameter datalogger
5-19
-------
Table 5-3
Data Archiving Formats
VEHICLE FILE (VEH.SSD)
Sample ID
VIN
Model Year
Make
Model
Transmission Type
Initial Odometer Reading
Load Measurement Type (MAP, MAF, or LV8)
Throttle Closed Position Voltage (mV)
Throttle Open Position Voltage (mV)
OPERATION FILE (OPN.SSD)
VIN
Date (DDMMMYY)
Local Time (HH:MM:SS)
Speed (miles/hour)
RPM (rpm)
MAP (kPa)
MAF (kg/hr)
LV8 (8 bit MAF)
Throttle Position (%)
Coolant Temperature (°C)
Equivalence Ratio
WIDE OPEN THROTTLE FILE (WOT.SSD)
VIN
Date (DDMMMYY)
Local Time (HH:MM:SS)
Speed (miles/hour)
RPM (rpm)
MAP (kPa)
MAF (kg/hr)
LV8 (8-bit MAF)
Throttle Position (%)
Coolant Temperature (°C)
Equivalence Ratio
5-20
-------
Table 5-3
(Continued)
ERROR FILE (ERR.SSD)
VIN
Local Time (DDMMMYY:HH:MM:SS)
Error Flag
Group
Logged Speed (m/s)
Logged RPM (rpm)
Logged MAP (kPa)
Logged MAF (kg/hr)
Logged LV8 (8-bit MAF)
Logged Throttle Position (%)
Logged Coolant Temperature (°C)
Logged Equivalence Ratio
Corrected Speed (m/s)
Corrected RPM (rpm)
Corrected MAP (kPa)
Corrected MAF (kg/hr)
Corrected LV8 (8-bit MAF)
Corrected Throttle Position (%)
Corrected Coolant Temperature (°C)
Corrected Equivalence Ratio
Record Deletion Flag
5-21
-------
installations for the purpose of converting the logged throttle position voltage to percent
throttle position.
The operation file (OPN.SSD) is the largest file of the four files and
contains 6,940,411 observations for the 216 vehicles. The parameters contained in this
file are shown in Table 5-3. The parameter called LV8 is General Motors' measurement
of MAP. It has values from 0 to 255 and is an 8-bit representation of relative mass
airflow.
The data collected on the 3-parameter datalogger during the installation
and removal drives by the study team were distinguished during data collection by a flag
in the raw data. Thus, the data present in the operations file for 3-parameter
dataloggers are only the data logged by the normal driver of the vehicle.
The 6-parameter dataloggers used a special procedure to avoid collecting
driving data made by the crew during installation and removal of the dataloggers. For
the GM datalogger, the operations data was erased just after the datalogger installation
and calibration check had been completed and just before the vehicle was released to
the owner for the one-week data collection period. At the end of the one-week period,
the removal crew removed the datalogger without driving the vehicle. For the non-GM
dataloggers, the installation drives were made with the memory modules disconnected.
Just before release of the vehicle to the driver, the datalogger internal memory was
erased of operations data and the empty memory modules were connected to the
datalogger. At the end of the one-week period, the memory modules were disconnected
without driving the vehicle.
The wide-open throttle file (WOT.SSD) contains the vehicle operation data
obtained during wide-open throttle accelerations of 3-parameter-instrumented vehicles.
For most of these vehicles, two wide-open throttles were recorded. The format of this
file is exactly the same as the format of the operation file. There is no wide-open
5-22
-------
throttle acceleration data for any of the 6-parameter instrumented vehicles, since the
6-parameter datalogging system was not able to distinguish between calibration driving
and vehicle owner driving.
The error file (ERR.SSD) is not used directly in the analysis of the data.
Instead, it is a documentation of the changes made to the raw data by the data quality
checking team. It contains the VIN and the time of each suspect group, which was made
up of each observation that was flagged and the surrounding observations. Each group
contains all of the parameters as they were logged and all of the parameters as they were
corrected during quality checking. When an entire parameter for a vehicle file was
changed to a missing value (for example, if all the MAPs were changed to missing
values), the corrected values are not part of the error file. Instead, this information is
given in Section 5.4 for all of the vehicles.
5.4 Results of Data Quality Checking
For each of the 293 data files produced by each of the instrumented
vehicles, the data quality checking methodology was used. At one point in that
methodology, the quality of the data was judged to be good or suspect for each vehicle
file. The results of this decision and the reason for labelling certain data files suspect
are given in Table 5-4. In addition, for a number of data files which were judged to be
good, some of the parameters were found to be unreliable and were therefore changed
to missing values for the entire data file. This is also shown in the table. Overall, the
table shows that of the 293 instrumented vehicles, 216 vehicles, or 74%, contained data
judged to be entirely reasonable.
The data analysis for this study was performed on data from the 216
vehicles. In the future, the data for the suspect vehicle data files will be re-examined to
determine if the data can be made good by judicious editing.
5-23
-------
Table 5-4
Summary of Data Quality Checking
Sample
Logger
Changed to Missing Values in Database
Number
Type
RPM
MAP
Coolant
Throttle
Equiv.
Reason for Labelling Suspect
Temp.
Position
Ratio
S 002
3
Good speed data
S003
6
X
Good speed data
S 004
3
Good speed data
S005
3
NA
NA
NA
NA
NA
No data in logger; RPM = 0 all the time.
S006
6
Good speed data
S007
6
X
Odd end of trips
S 008
3
Good speed data
S01 1
3
Good speed data
S01 4
6
X
Good speed data
S01 5
3
Good speed data
S01 8
3
Good speed data
S01 9
6
X
Good speed data
S024
3
Good speed data
S 025
3
NA
NA
NA
NA
NA
Uncalibrated datalogger
S 027
3
X
Good speed data
S028
6
X
Good speed data
S030
3
NA
NA
NA
NA
NA
Datalogger shut off in mid-trip
S036
3
Good speed data
S038
3
Good speed data
S 040
6
X
Good speed data
S041
3
Good speed data
S 042
6
NA
NA
NA
NA
NA
Noise in speed at idle
S 043
3
NA
NA
NA
NA
NA
Numerous speed spikes
S 045
6
X
Good speed data
S 046
3
Good speed data
S047
6
Good speed data
S 048
3
Good speed data
S050
3
Good speed data
S051
6
X
Good speed data
S053
6
Good speed data
S054
3
Good speed data
S055
3
Good speed data
S058
6
NA
NA
NA
NA
NA
Noise in speed at idle
S061
3
Good speed data
S 063
6
X
Good speed data
S 066
3
NA
NA
NA
NA
NA
Logger shutoff during trip
S067
3
Good speed data
S069
3
Good speed data
S070
6
Good speed data
S072
3
Good speed data
S073
3
Good speed data
S076
3
NA
NA
NA
NA
NA
Numerous speed spikes. Engine misfiring
S078
3
Good speed data
S079
3
NA
NA
NA
NA
NA
Speed fluctuates in an uncorrectable manner
S082
6
NA
NA
NA
NA
NA
Noise in speed at idle
S088
3
Good speed data
S 089
6
Good speed data
S 090
3
Good speed data
S092
6
Good speed data
S093
6
Good speed data
5-24
-------
Table 5-4
(Continued)
Sample
Number
Logger
Type
Changed to Missing Values in Database
Reason for Labelling Suspect
RPM
MAP
Coolant
Temp.
Throttle
Position
Equiv.
Ratio
S094
3
Good speed data
S096
3
NA
NA
NA
NA
NA
Lost speed calibration
S097
6
Good speed data
S098
6
X
Good speed data
S1 00
3
Good speed data
S1 03
3
Good speed data
S1 05
3
Good speed data
S1 07
6
Good speed data
S1 08
3
NA
NA
NA
NA
NA
Memory turning off in mid-trip
S109
6
X
Good speed data
S1 10
3
NA
NA
NA
NA
NA
Speed fluctuation. Lost speed calibration
S1 12
3
Good speed data
S1 14
3
Good speed data
S1 17
3
Good speed data
S1 18
3
Good speed data
S1 19
6
X
Good speed data
S120
3
Good speed data
S122
3
NA
NA
NA
NA
NA
Shuts off in mid-trip. Speed sensitivity high
S1 25
6
NA
NA
NA
NA
NA
Noise in speed at idle
S1 26
3
Good speed data
S127
6
X
Turns off in mid-trip
S1 28
3
Good speed data
S1 29
6
X
Good speed data
S1 30
3
NA
NA
NA
NA
NA
Turns off in mid-trip. Holes in speed data
S1 31
6
X
X
X
X
Good speed data
S1 32
3
NA
NA
NA
NA
NA
Many speed problems
S1 33
6
Good speed data
S1 34
3
X
Good speed data
S1 35
3
Good speed data
S1 37
3
Good speed data
S1 38
3
Good speed data
S1 43
3
Good speed data
S1 44
3
Good speed data
S146
3
NA
NA
NA
NA
NA
Holes in speed data. Turns off in mid-trip
S1 47
3
Good speed data
S1 48
6
Good speed data
S1 50
6
NA
NA
NA
NA
NA
Noise in speed at idle
S1 52
3
NA
NA
NA
NA
NA
Logger turns off every five seconds
S1 54
3
NA
NA
NA
NA
NA
Noise in speed at idle
S156
3
Good speed data
S1 58
3
Good speed data
S160
3
NA
NA
NA
NA
NA
Noise in speed at idle
S1 63
3
Good speed data
S1 64
6
X
Good speed data
S1 65
3
Good speed data
S1 67
3
Good speed data
S1 69
3
Good speed data
S1 72
3
Good speed data
S1 75
3
Good speed data
S1 76
3
Good speed data
5-25
-------
Table 5-4
(Continued)
Sample
Logger
Changed to Missing Values in Database
Number
Type
RPM
MAP
Coolant
Throttle
Equiv.
Reason for Labelling Suspect
Temp.
Position
Ratio
S 1 77
6
Good speed data
S 1 78
3
Good speed data
S 1 79
3
Good speed data
S 1 80
3
Good speed data
S1 81
6
X
Good speed data
S1 34
3
Good speed data
S 1 87
3
Good speed data
S1 89
3
NA
NA
NA
NA
NA
Logger shut off during 2 trips
S1 94
3
Good speed data
S196
3
NA
NA
NA
NA
NA
Logger shutdown from RPM drop below 300
S197
3
Good speed data
S200
3
Good speed data
S201
6
X
Good speed data
S203
3
Good speed data
S204
3
NA
NA
NA
NA
NA
Noise in speed at idle
S 205
3
Good speed data
S 206
6
NA
NA
NA
NA
NA
Noise in speed at idle
S207
3
NA
NA
NA
NA
NA
Logger turns off in mid-trip; Noise in speed.
S 209
3
Good speed data
S21 0
6
NA
NA
NA
NA
NA
Speed sensor went bad at -11000th sec.
S211
3
Good speed data
S212
3
Good speed data
S214
3
Good speed data
S215
6
NA
NA
NA
NA
NA
No disk
S 21 7
6
X
Good speed data
S218
3
Good speed data
S219
3
NA
Good speed data
S220
6
NA
NA
NA
NA
Noise in speed at idle
S221
3
Good speed data
S 222
3
Good speed data
S224
3
NA
Good speed data
S 226
3
NA
NA
NA
NA
Numerous speed spikes
S227
3
Good speed data
S229
3
NA
Good speed data
S233
3
NA
NA
NA
NA
Noise in speed at idle
S235
3
Good speed data
S236
6
Good speed data
S238
3
NA
NA
NA
NA
NA
Noise in speed at idle
S239
3
Good speed data
S240
3
NA
NA
NA
NA
NA
Noise in speed input at all speeds
S241
3
Good speed data
S243
6
Good speed data
S 244
3
NA
NA
NA
NA
NA
Noise in speed at idle
S246
3
NA
NA
NA
NA
NA
Noise in speed at idle
B002
3
Good speed data
B005
6
Good speed data
B007
3
NA
NA
NA
NA
NA
Noise in speed
B008
6
Good speed data
B010
3
Good speed data
B01 1
6
Good speed data
5-26
-------
Table 5-4
(Continued)
Sample
Logger
Changed to Missing Values in Database
Number
Type
RPM
MAP
Coolant
Throttle
Equiv.
Reason for Labelling Suspect
Temp.
Position
Ratio
B016
3
Good speed data
B017
3
Good speed data
B019
3
Good speed data
B037
3
Good speed data
B039
3
Good speed data
B042
3
Good speed data
B044
6
X
Good speed data
B 048
3
NA
NA
NA
NA
NA
Logger shut off a lot
B050
3
Good speed data
B052
3
Good speed data
B053
6
Good speed data
B055
6
X
Good speed data
B058
3
Good speed data
B059
3
Good speed data
B062
3
Good speed data
B065
6
NA
NA
NA
NA
NA
Noise In speed at idle
B066
3
Good speed data
B072
3
Good speed data
B074
6
Good speed data
B076
6
NA
NA
NA
NA
NA
Noise in speed at idle
B079
3
NA
NA
NA
NA
NA
Noise in speed
B081
3
Good speed data
B084
3
Good speed data
B086
3
Good speed data
B088
3
Good speed data
B098
3
X
Good speed data
B099
6
X
Good speed data
B101
6
X
Logger shuts off
B1 04
3
NA
NA
NA
NA
NA
No data on disk
B1 05
3
Good speed data
B1 10
3
NA
NA
NA
NA
NA
No data on disk
B11 3
6
X
Good speed data
B117
3
NA
NA
NA
NA
NA
No data on disk
B1 19
6
X
Good speed data
B123
3
Good speed data
B1 24
3
X
Good speed data
B130
3
Good speed data
B1 39
3
Good speed data
B1 40
6
Good speed data
B143
3
X
Good speed data
B144
6
Good speed data
B1 45
6
Logger shuts off
B150
3
Good speed data
B153
3
Good speed data
B1 56
3
Good speed data
B159
6
X
Logger shuts off
B 1 63
3
X
Good speed data
B 1 64
3
Good speed data
B1 65
6
Good speed data
B 1 67
6
X
Good speed data
5-27
-------
Table 5-4
(Continued)
Sample
Logger
Changed to Missing Values in Database
Number
Type
RPM
MAP
Coolant
Throttle
Equiv.
Reason for Labelling Suspect
Temp.
Position
Ratio
B 1 69
3
Good speed data
B 1 78
3
X
Good speed data
B 1 84
3
Good speed data
B 1 90
3
X
Good speed data
B 1 91
3
Good speed data
B1 93
3
Good speed data
B 1 95
3
Good speed data
B 1 96
6
Logger shuts off
B 2 06
3
Good speed data
B207
3
NA
NA
NA
NA
NA
Noise in speed
B208
3
X
Good speed data
B210
3
Good speed data
B217
3
NA
NA
NA
NA
NA
Noise in speed
B233
3
X
Good speed data
B236
3
Good speed data
B 238
6
NA
NA
NA
NA
NA
No disk
B241
3
NA
NA
NA
NA
NA
Noise in speed
B243
3
Good speed data
B247
3
Good speed data
B 2 55
3
Good speed data
B256
3
NA
NA
NA
NA
NA
0.8g accels unbelievable
B263
3
NA
NA
NA
NA
NA
Noisy speed. Excursions to 40.95m/s
B268
3
X
Good speed data
B269
3
NA
NA
NA
NA
NA
Abrupt high speed at trip starts.
B273
3
Good speed data
B275
3
X
Good speed data
B 280
3
NA
NA
NA
NA
NA
Noisy speed
B281
6
NA
NA
NA
NA
NA
Noise in speed at idle
B282
6
NA
NA
NA
NA
NA
Noise in speed at idle
B 286
3
X
Good speed data
B287
3
Good speed data
B291
6
NA
NA
NA
NA
NA
Noise in speed at idle
B293
3
NA
NA
NA
NA
NA
Speed Bad
B297
3
Good speed data
B301
3
NA
NA
NA
NA
NA
Spikes in speed
B302
6
X
Good speed data
B314
3
Good speed data
B315
3
Good speed data
B317
3
Good speed data
B319
3
Good speed data
B324
6
NA
NA
NA
NA
NA
Noise in speed at idle
B325
3
NA
NA
NA
NA
NA
Parity errors
B329
3
Good speed data
B337
3
Good speed data
B338
3
X
Good speed data
B344
3
Good speed data
B350
3
Spikes & noise in speed
B351
3
Good speed data
B354
6
Good speed data
B358
6
Good speed data
5-28
-------
Table 5-4
(Continued)
Sample
Logger
Changed to Missing Values in Database
Number
Type
RPM
MAP
Coolant
Temp.
Throttle
Position
Equiv.
Ratio
Reason for Labelling Suspect
B 361
3
X
Good speed data
B363
6
Good speed data
B365
3
X
Good speed data
B367
3
Good speed data
B368
3
X
Good speed data
B369
3
X
Good speed data
B370
6
X
Good speed data
B 371
3
Good speed data
B375
3
Good speed data
B376
3
Good speed data
B381
3
X
Good speed data
B386
3
X
Good speed data
B 389
3
Good speed data
B 390
6
No disk
B392
6
Good speed data
B395
3
Good speed data
B398
3
No speed calibration
B406
3
Good speed data
B410
3
Good speed data
B413
3
Good speed data
B414
6
X
Logger shuts off
B419
3
Good speed data
B 420
3
X
Good speed data
B424
3
NA
NA
NA
NA
NA
Speed noise
B426
3
NA
NA
NA
NA
NA
Very jerky speed transducer
B 428
3
NA
NA
NA
NA
NA
Speed spiking
B431
6
X
Logger shuts off
B433
6
X
Logger shuts off
B435
6
Good speed data
B436
3
X
Good speed data
B438
3
Good speed data
B441
3
Good speed data
B445
3
X
Good speed data
B447
3
X
Good speed data
B451
3
Good speed data
B452
3
NA
NA
NA
NA
NA
Noisy speed for a period of time
B454
6
Good speed data
B460
3
Good speed data
B466
3
NA
NA
NA
NA
NA
No speed reading
B467
3
X
Good speed data
B468
3
Good speed data
B472
3
Good speed data
B482
3
NA
NA
NA
NA
NA
Noisy speed at high speeds
5-29
-------
6.0 DATA ANALYSIS
The immediate objective of this project is to collect driving pattern data to
evaluate the representativeness of the current FTP driving cycle. If it is found that the
current FTP cycle does not accurately represent today's driving, then a new test
procedure may need to be developed. Although the development of a new driving cycle
is not part of this work, it is useful to consider the methodology that might be used to
evaluate or develop candidate driving cycles, because it will help guide analyses of the
driving pattern data and make development or evaluation of candidate driving cycles
more efficient.
There appear to be at least two ways to arrive at new driving cycles for a
new Federal Test Procedure. The more conventional of the two methods is to select a
driving pattern experienced by an actual car in the study that is believed to be
representative of all of the driving done in the study. The second method is to use all of
the driving data in the database to develop a procedure for simulating driving patterns.
For both approaches, measures of the representativeness of the chosen cycle or cycles
need to be determined.
Representativeness can be measured by calculating statistics on the driving
cycle candidate and calculating the same statistics on all of the data in the database. If
all of the statistics that can possibly be imagined agree exactly for both the driving cycle
candidate and the database, the driving cycle is said to be representative of the driving.
In practice, it is unlikely that a driving cycle of reasonable length will exactly match any
and all possible statistics of the database driving behavior data. Instead, a more practical
goal is to find driving cycle candidates that match the statistics of the database as closely
as possible.
The driving behavior of vehicles is made up of two basic parts: soaks and
trips. Both of these features need to be considered when evaluating the driving pattern
6-1
-------
data. Soaks are important for both exhaust and evaporative emissions because of the
thermal nature of today's vehicle propulsion systems. The length of the soak influences
the exhaust emissions of the following trip. For evaporative emissions, the temperature
level of the vehicle power plant during soaks and the length of the soak periods are
important factors. When the soaks occur is not as important as the length of the soaks
between the trips.
The detailed speed behavior within a trip is also an important
characteristic of driving behavior. The reason for this is that the emissions behavior of
today's vehicles depend not only on the fact that the engine is on, but on the loads and
the duration of the loads that the engine experiences.
Because statistics will be used to determine whether a candidate driving
cycle is a good representative of the driving behavior of today's vehicles, the first section
contains the results of many analyses of the database for reference purposes. Analyses
are performed for speed and acceleration measures, measures of trips, and vehicle/driver
measures. This statistical analysis should be considered to be preliminary in nature
because of the size of the database and the short time devoted to the task. A review of
the results of the statistical analyses will lead any investigator to ask many more
questions and to desire to conduct additional analyses of the database.
The same statistics are calculated for the FTP cycles for comparison with
the statistics on the driving pattern database.
A review of biases which may be present in the data is reviewed in the next
section. Because of the algorithms that were used to calculate speed in the datalogger
and acceleration in the data analysis phase, certain biases may exist. This discussion
provides an estimate of the magnitude of these biases. While attempts were made to
select vehicles in a representative manner for the 3-parameter datalogger installations, it
is necessary to perform statistical tests to determine if the sample taken was
6-2
-------
representative of the population. Finally, if the driver of a vehicle knows that a
datalogger is collecting data from his vehicle, it is possible that the driver may alter his
behavior. This is examined in the data bias section by looking at driving behavior at the
beginning of the week, in comparison with driving behavior at the end of the week.
The next section presents two options for advanced analysis and simulation
of driving patterns. The two options are based on advanced statistical analysis and on
signal processing approaches. For both options, the emphasis is on the selection or
development of driving cycles representative of the driving patterns measured in this
study. Neither of these approaches is implemented in this study, but they serve as ideas
for the development of new driving cycles and test procedures.
Finally, in the last section, options for formulating driving cycles are
presented. The EPA need not necessarily concentrate on developing a Federal Test
Procedure that involves a single cast-in-iron driving cycle for the next generation of
certification procedures. Several other options can be considered, and each has its
advantages and disadvantages.
6.1 Statistical Analysis
The summary statistics for trips and soaks are presented in this section. To
perform the routine statistical analysis of the driving patterns database, a series of
definitions and description of variables are presented. In later sections, the results of
speed and acceleration measures, trip measures, and driver and vehicle measures are
presented. Finally, the same types of measures for the current Federal Test Procedure
are calculated and compared with the statistics that were calculated on the driving
pattern database.
Because many different statistics have been calculated, they are presented
in different ways. Many of the statistics are described using distribution plots in the text.
-------
In many cases, these are plotted on a log scale because of the large range the
distributions span. As in any technical area, distribution plots made on linear or log
scales require some understanding of how these results are to be interpreted. Care
should be exercised when using these distribution plots to arrive at conclusions about
trends in the data. Additional aids to the reader are provided by bar charts and tabular
listings of the values used for the bar charts and log plots. These aids are in Appendix
B. Finally, the averages of many of the parameters are provided in tables in the
conventional format of mean, standard deviation, minimum value, and maximum value.
The values calculated from the driving patterns database were compared
whenever possible with the results obtained in the 1979 General Motors driving survey,
in which GM had NPD Research, Inc. (5) conduct a diary survey nationwide of about
2000 households. Drivers recorded the trip start and trip end data in a notebook for
every trip that they made during a one-week period. About 73,000 trips were recorded
for 3157 vehicles. The two studies, however, may yield slightly different results because
the methods used to acquire the data were different. In this driving patterns study,
electronic dataloggers were used, whereas in the General Motors survey written survey
instruments were used to record information.
6.1.1 Definitions of Terms
Table 6-1 lists the definitions of the terms used in this analysis. The
definition of trip needs some discussion, because its meaning must be known to allow
interpretation of the results of the statistical analysis.
Trip has been defined in Table 6-1 to have the same meaning as the period
during which the dataloggers turned on and off during data collection periods on the
vehicles. In general, this beginning and end of trip was caused by the engine being
turned on and off. However, in the normal use of a vehicle, the actual engine on and
6-4
-------
Table 6-1 Definition of Terms
Idle: The mode when the engine is on and the vehicle has a speed of 0.00
mph.
Running: The mode when the engine is on and the vehicle speed is
greater than 0.00 mph.
Trip: A trip starts every time the engine is turned on and ends when the
engine is turned off.
Beginning of a Stop: Occurs when vehicle speed first drops below 4 mph
after having been above 10 mph and speed has not already been below 4
mph since speed dropped below 10 mph.
End of a Stop: Occurs when vehicle speed first exceeds 4 mph if speed
eventually goes above 10 mph without first dropping back below 4 mph.
Beginning of a Creep: Occurs when vehicle speed first exceeds 4 mph and
remains below 10 mph and is followed by the end of a creep.
End of Creep: Occurs when vehicle speed drops below 4 mph after having
been below 10 mph since the speed was above 4 mph.
Observation Phase (transition, stable): Transition phase extends from
logger installation to the second 1:00 am after installation. The stable
phase extends from the second 1:00 am after installation to the removal of
the logger.
MPH (mph): Vehicle speed averaged over a one second period.
Resolution is 0.02 mph for the 3-parameter datalogger. Resolution varies
for the 6-parameter datalogger. For example, resolution is 1 mph for the
GM datalogging system.
ACCEL (mph/s): Average vehicle acceleration from the previous second to
this second.
POWER (mph2/s): Average Specific Power from the previous second to
this second.
= if speed(t) £ speed(t-l)
= speed2(t) - speed2(t-l), if speed(t) > speed(t-l).
6-5
-------
engine off events may not be those which were intended by the driver. That is, the
engine may do something different than what the driver wants it to do.
The most common example of this is starting an engine after a long soak.
The driver intends for the engine to start, to idle a short period, and for the vehicle to
begin moving. However, because the engine is cold, it may not start on the first attempt.
Therefore, there may be one or more periods when the engine starts and dies. This stall
can be repeated several times. According to the trip definition, a stall is a trip; it is
followed by a short soak.
To evaluate the effects on the statistical results of defining a trip in one
way or another way, EPA requested that an attempt be made to find a way to identify
very short duration trips which are adjacent to long duration trips. The concept here is
that if the vehicle had started properly the first time, the intent of the driver would have
been realized, and a fewer number of trips would have occurred. The problem in trying
to identify these minor trips is that the datalogger has no way of knowing the intent of
the driver. Perhaps he turned off the engine immediately after starting it to perform
some other very short task. In addition, the vehicle emissions do not respond to the
driver's intent.
An analysis of the 3-parameter data taken in Spokane was examined to
determine if there was a way to separate the minor trips from the remainder of the trips,
which will be called major trips. Two frequency distributions were created to look for
this separation. In the first, the frequency of the trip duration versus the duration of
soak after the trip was created. This frequency distribution showed an island of
observations for short trips with short post-soaks. A natural valley in the data was
observed for trip durations at about 18 seconds and at post-soak duration of about 18
seconds. These minor trips would include those in which a cold engine had stalled
during a start, but then was started again a short time later. Approximately 7% of the
Spokane trips were in this island. Another frequency distribution was made of trip
6-6
-------
duration versus the soak time before the trip. In this distribution, a smaller number of
trips were found to have short durations and short soaks preceding them. These short
trips would not be expected to be caused by stalls at cold starts, because they had
occurred after the engine had been running. A natural valley falls in the data for
durations less than 6 seconds and presoak times less than 10 seconds. In this small
island, only about 1.4% of the Spokane trips were found.
Because the presence of a large number of short duration trips can affect
the statistics on the trips in the entire data set, it is desirable to calculate the trip
measures and vehicle and driver measures, both with major and minor trips (that is, all
of the trips) and with major trips alone (that is, leaving out the minor trips). For this
analysis, the minor trips which have been left out of one analysis are those which have
durations less than 18 seconds with a following soak of less than 18 seconds.
Table 6-2 lists the computed variables and their descriptions. These
variables are used throughout this section to calculate statistics on the driving pattern
data.
6.12 Discussion of Results
The goal of this section on data analysis is to evaluate the driving pattern
data and summarize vehicle driving behavior in Spokane and Baltimore. The analysis
was directed to consider three different aspects of vehicle driving: speed and
acceleration measures, trip measures, and vehicle/driver measures. At EPA's request,
summary statistics were developed for each of these categories. The statistics for all
observations in the data set are presented in Table 6-3; these are discussed later at
several places in this section. Speed and acceleration measures summarize the second-
by-second driving behavior reflected by vehicle speed and acceleration. The trip
measures describe patterns of time and distance spent on individual trips. The
vehicle/driver measures describe differences in operating individual vehicles.
6-7
-------
Table 6-2 Variable Descriptions
SITE: Denotes the I/M station where vehicle was solicited: Spokane,
Baltimore-Rossville, or Baltimore-Exeter.
SAMPLE: Sample identification number unique to each solicited vehicle
whether it was instrumented or not.
VIN: Vehicle identification number.
MAKE: Make of the vehicle.
MODEL: Model of the vehicle.
YEAR: Model year of the vehicle.
TPL: Throttle Closed Position Voltage (mV) for 6-parameter dataloggers.
TPH: Throttle Open Position Voltage (mV) for 6-parameter dataloggers.
TR: Transmission type: A for automatic, M for manual.
SP5: Speed classification with a width of 5 mph. For example, SP5 = 15
contains observations with: 10 < speed (mph) £15. For idle, SP5 = 0
contains only observations with speed (mph) = 0.
AC1: Acceleration classification with a width of 1 mph/s. For example,
AC = -3 contains observations with: -4 < acceleration (mph/s) £ -3.
TRIPTIME (s): The duration of a trip from engine on to engine off.
T IDLE (s): The time spent in the idle mode.
TJDLE + T_RUN = TRIPTIME.
T FIDLE (s): The time spent in idle after start and before the vehicle
moves for a trip.
T RUN (s): The time spent in a trip at non-zero speed.
AVG_MPH (mph): Average speed for a trip.
6-8
-------
Table 6-2 continued
SOAKTIME (s): Soak time before a trip begins.
TRIPDIST (mile): The distance travelled during a trip.
A_STOPD (mile): Average distance between stops for a trip.
AV_R_MPH (mph): Average speed during the non-zero speed portion of
a trip.
AVE_PKE (mph2/s): Average specific power for the trip.
VEHMIDAY (minutes): Average vehicle operation time per day for a
vehicle.
VEHDIDAY (miles): Average distance driven per day for a vehicle.
TRIPSDAY: Average number of trips driven per day for a vehicle.
NUMSTHR: Average number of stops per hour of use for a vehicle.
ACCGT3 (%): Percent of observations with accelerations £ 3 mph/s.
Percent of observations with accelerations £ 4 mph/s.
Percent of observations with accelerations £ 5 mph/s.
Percent of observations with accelerations S 6 mph/s.
Percent of observations with speeds £ 0 mph.
Percent of observations with speeds £ 10 mph.
Percent of observations with speeds £ 20 mph.
Percent of observations with speeds £ 30 mph.
Percent of observations with speeds £ 40 mph.
Percent of observations with speeds £ 50 mph.
Percent of observations with speeds £ 60 mph.
ACCGT4 (%):
ACCGT5 (%):
ACCGT6 (%):
MPHGTO (%):
MPHGT10 (%):
MPHGT20 (%):
MPHGT30 (%):
MPHGT40 (%):
MPHGT50 (%):
MPHGT60 (%):
-------
Table 6-2 continued
MPHGT70
(%):
Percent
of
observations
with
speeds > 70 mph.
MPHGT80
(%):
Percent
of
observations
with
speeds > 80 mph.
PKEGT100
(%):
Percent
of
observations
with
specific
power
>
100
mph2/s
PKEGT120
(%):
Percent
of
observations
with
specific
power
>
120
mph2/s
PKEGT140
(%):
Percent
of
observations
with
specific
power
>
140
mph2/s
PKEGT160
(*>):
Percent
of
observations
with
specific
power
>
160
mph2/s
PKEGT180
(%):
Percent
of
observations
with
specific
power
>
180
mph2/s
PKEGT200
(%):
Percent
of
observations
with
specific
power
>
200
mph2/s
PKEGT220
(%):
Percent
of
observations
with
specific
power
>
220
mph2/s
PKEGT240
(%)¦
Percent
of
observations
with
specific
power
>
240
mph2/s
PKEGT260
(%):
Percent
of
observations
with
specific
power
>
260
mph2/s
PKEGT280
(%):
Percent
of
observations
with
specific
power
>
280
mph2/s
PKEGT300
(%):
Percent
of
observations
with
specific
power
>
300
mph2/s
6-10
-------
Table 6-3
Summary of Statistics for All Observations
in the Driving Pattern Data Set
Variable
Units
Number of
Weighted by
Mean
Standard
Minimum
Maximum
Observations
Deviation
M=H
(mph)
6927206
None
24.70
20.04
O.OO
94.46
Acca
(mph/s)
6915785
None
0.00
1.51
-33.20
33.34
PCWER
(mpb*2/s)
2690251
None
48.83
48.29
0 oo
2876.14
TRIPTIME
(*)
12073
None
575
650
1
14797
T IDLE
(s)
1 1 532
None
112
173
0
3182
T FIDLE
(5)
12073
None
28
81
0
2085
T RUN
(s)
11 532
None
515
573
0
14768
AVE_MPH
(mph]
11 405
None
18.18
10.78
0.00
71.36
SOAKTIME
(s)
11 857
None
11 719
24075
2
416053
TRIPDIST
(miles)
12073
None
3.94
7.50
0.00
255.00
A_STOPD
(miles)
10300
None
0.88
2.24
0.00
127.48
AV R MPH
(mph)
10817
None
23.11
10.33
0.02
71.63
AVE_PKE
-------
6.1.2.1
Speed and Acceleration Measures
As part of this analysis, the second-by-second speed and acceleration
behavior of 216 vehicles were evaluated to consider the range of behavior of the overall
vehicle population. Table B-l lists these vehicles. Descriptive statistics for speed and
acceleration, as well as the time spent at any given speed or acceleration, are presented
in this section.
Figure 6-1 shows a contour plot of speed versus acceleration for 216
vehicles instrumented in both Spokane and Baltimore. This contour plot was developed
by using speed classes of 2 mph and acceleration classes of 0.5 mph/s. This figure
represents 6.9 million observations. The shading of each cell in the contour plot
represents the percent of all observations that satisfy the boundary limits of the cell. In
general, the figure shows that a large percentage of vehicle operation time was spent at
idle. Large peaks are seen around 35 mph and around 60 mph. A higher percentage of
positive accelerations is seen between 0 and 25 mph. This can be attributed to similar
acceleration behavior for drivers while they have dissimilar deceleration behavior. In
addition, it shows that most of the vehicle operation is over mild acceleration conditions
between -2 and +2 mph/s. However, accelerations as high as 6.0 mph/s were observed,
and many accelerations above 3.3 mph/s, the FTP maximum acceleration, were common.
Table B-2 shows the percent of vehicle operation time spent at any given speed and
acceleration condition.
Figure 6-2 shows the distribution of speeds for the vehicles in the two
cities. 19.5% of the vehicle operation time was spent while the vehicle speed was 0 mph.
Consistent with the contour plot, about 11% of the vehicle operation time was spent
between 30 and 35 mph; in addition, a smaller peak occurs between 55 and 60 mph.
Thus, while drivers drove their vehicles at many speeds, 0, 35, and 60 mph stand out as
being most common. Figure 6-3 shows the range of accelerations for the driving pattern
data. The figure shows the occurrence of a high percentage of 0 and near 0
6-12
-------
Figure 6-1. Speed/Acceleration Contours for All Observations
-10
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
Speed (mph)
Percent of i i
Observations 0.20 mmw 0.50
0.01 T77r77\ o.02
POOOO 0.05
B 1.00
'////////.
0.10
2.00
-------
Figure 6-2. Speed Distribution for All Observations
Os
20-
19-
18 -
V)
r
17-
O
16-
•
15-
o
14 -
>
13 -
-------
Figure 6-3. Acceleration Distribution for All Observations
Acceleration (mph/s)
-------
accelerations, but the detailed values show that 3.75% of the acceleration observations
are below -3.3 mph/s, and about 2.67% of the acceleration observations were above 3.3
mph/s. Tables B-3 and B-4 present the speed and acceleration information in tabular
form.
Figure 6-4 shows the specific power distribution in classes of 20 mph2/s.
Table B-5 presents the same information in tabular form. The term, specific power, in
this context relates to the increase in kinetic energy per second per unit mass. As can be
seen from the figure and the table, 99% of the observations had specific powers below
240 mph2/s. The percent of observations as a function of specific power was found to
closely follow an exponential decay function. The cumulative percent of observations
can be represented by the following equation:
Q = 100 * (l-ek"p)
where
k is the decay coefficient = 0.020 s/mph2
P = specific power in mph2/s
For example, by using the above equation, the cumulative percent of specific power
values up to 100 mph"/s is calculated to be 86.47 percent. The corresponding measured
value given in Table B-5 is 87.2 percent. This equation can also be used to compare the
percent of observations between two specific power values. Because the cumulative
percent of observations is modeled, the width of the classes used for the specific power
distribution does not affect the equation. For example, the percent of observations
between 80 and 100 mph2/s is calculated to be 6.66% by using the equation twice. The
measured value is 7.09% based on Table B-5.
Table 6-3A presents the mean, standard deviation, minimum, maximum,
and count of seconds for speed, acceleration, and specific power. As shown, the average
speed for the 216 vehicles was 24.7 mph, with a standard deviation of 20.0 mph.
6-16
-------
Figure 6-4. Specific Power Distribution for All Observations
On
I
--4
40 A
w
c
o
•-Z3 30
O
>
i n i i t i i
c
(D
u
aj
Q_
(U
>
40
O
Z5
30 E
20
10
0
0 50 100 150 200 250 300 350 400 450 500
Specific Power (mph**2/s)
-------
6.1.2.2
Trip Measures
As mentioned earlier, a trip was defined for this preliminary analysis as the
vehicle operating time between an engine start and an engine shutdown. Because of this
definition of a trip, if a vehicle's engine had false starts, each may be counted as a trip.
Figure 6-5 shows a distribution of 12,073 trip times for all trips in both Spokane and
Baltimore for the entire driving modes data set. About 10% of the trips were shorter
than 15 seconds, and are shown by the vertical bar between 0.2 to 0.25 minutes. The
rest of the trips show a smooth distribution which is close to log normal.
The minor trips have been removed to produce the distribution of trip
times in Figure 6-6. The median trip length for the trip time distribution is about 7
minutes. The GM median trip time was 10.0 minutes. It is unlikely that very short trips,
such as engine false starts, were recorded in the diary survey. Table B-8 shows the trip
time distribution in tabular form.
Figure 6-7 and Table B-ll show the soak time distribution for the driving
pattern data of all trips, and Figure 6-8 shows the soak time distribution for only the
major trips. Soak time is defined as the time between the trips. This can be a useful
measure for determining whether the vehicle was in a cold start mode during the start of
the following trip. This information is also very useful for determining evaporative
emissions. Both Figures 6-55 and 6-155 show a significant amount of short soak times,
less than 0.5 minutes in duration. Between about 3 minutes and 100 minutes, the
percent of soak periods remains constant at around 6 percent. A large peak of soak
periods occurs between about 600 and 1000 minutes. This relates to about 10 to 16
hours and corresponds to the overnight soak after a vehicle has been parked in the
evening and then switched on the following morning.
6-18
-------
Figure 6-5. Trip Time Distribution for AH Trips
OS
I
VO
1 1 ¦
10-
9-
cn
8
CL
L_
1
7-
• i
o
6-
-+—>
d
(D
5-
O
4-
|
2.0
' i t » » i r f |
20.0
' l l f
100
90
80
70
60
50
c
a>
o
o
Q_
40 O
3
30 E
D
20 O
7
200.0
10
I- 0
Trip Time (minutes)
-------
Figure 6-6. Trip Time Distribution for M^jor Trips Only
10-1
9
8
l t t * t I f I |
0.2 2.0
GM
Median
' i i i » t > i |
20.0
Trip Time (minutes)
100
90
80
\r 70
60
50 §
40 O
3
30 E
13
20 O
10
T
200.0
0
-------
8
7
6
5
4
3
2
1
0
Figure 6-7. Soak Time Distribution for All Trips
| * >f»f|M| I I f f I 1 111 > » > I MH|
.1 1.0 10.0 100.0
h 100
90
80
70 o
60
[ I f I 1 T H»|'
1000.0 10000.0
0
Soak Time (minutes)
-------
Figure 6-8. Soak Time Distribution for Mqjor Trips Only
Soak Time (minutes)
-------
Figures 6-9 and 6-10 show the time spent at idle and running time
distributions for the driving mode data. Both of these plots are similar to the trip time
distribution and are close to log normal distributions. From these plots, we can say that
for the statistical median trip of about 7 minutes, about 1 minute was spent at idle and
about 6 minutes were spent at a non-zero speed. Tables B-14 and B-17 show this
information in tabular form. Figures 6-11 and 6-12 show the same distributions for
major trips only; the distributions are very similar to the figures for all trips.
Figure 6-13 and Table B-20 show the distribution of trip distance in miles
for the driving modes data. Information for all 12,073 trips is shown in the figure and
the table. Again, the short trips appear to be a significant percentage, and the rest of
the plot shows a smooth single mode distribution. Figure 6-14 was obtained after minor
trips were removed. The median trip distance shown by the distribution is about 2.3
miles, and the median trip distance from the GM survey was about 3 miles. In addition,
the average distance for the driving modes data is about 3.9 miles, while the GM average
distance was 7.1 miles.
Table 6-3B shows the unweighted trip measures for the driving modes data
for all trips. Various trip measures were calculated separately for each trip. Table 6-3B
presents the average and the range of variation of these trip measures. For example, the
average of the speed observations from Table 6-3A for the entire data set was 24.7 mph,
but from Table 6-3B the average trip speed is 18.2 mph. This is because a 2-second trip
or 200-minute trip are treated equally to calculate the trip measures.
Table 6-3C shows that the average distance between stops for all the trips
weighted by the number of stops was about 0,83 miles, which can be compared to 0.88
miles in Table 6-3B. This suggests that if the number of stops were high for a particular
trip, the average distance between them was smaller, compared to trips with a fewer
number of stops.
6-23
-------
Figure 6-9. Time in Idle Distribution for All Trips
CT\
I
N>
100
0.04
0.40
4.00
Time in Idle (minutes)
-------
Figure 6-10. Running Time Distribution for All Trips
Running Time (minutes)
-------
Figure 6-11. Time in Idle Distribution for Major Trips Only
Total Time in Idle during a Trip(minutes)
-------
Figure 6-12. Running Time Distribution for Major Trips Only
Running Time during a Trip(minutes)
-------
Figure 6-13. Trip Distance Distribution for All Trips
o\
I
to
oo
0.01
0.10 1.00 10.00
Trip Distance (miles)
100.00
-------
Figure 6-14. Trip Distance Distribution for M^jor Trips Only
Trip Distance (miles)
-------
Table 6-3D shows the average means and standard deviations for average
speed, average running speed, and average specific power weighted by the trip distance.
As shown, the weighted average speed jumps to 30.7 mph from an unweighted average of
18.2. Table 6-3E shows the three descriptors, but this time weighted by trip time. The
weighted measures show that longer trips have higher average speeds, higher average
running speeds, and higher average specific power than shorter trips.
6.1.2.3 Vehicle/Driver Measures
In this section, measures that describe the operation of individual vehicles
are discussed. Vehicles were instrumented for about a week in the two cities. This
section presents the hourly or daily averages of vehicle operation measures.
Figure 6-15 and Table B-23 present the average distance that each vehicle
drove daily. This figure represents the driving information from 216 vehicles. It was
determined by first computing the total distance that each vehicle was driven for the
instrumentation period and the total number of days that each vehicle was instrumented.
The number of days was calculated by considering the time of start of the first trip and
the end of the last trip while the vehicle was instrumented. This time was then divided
by 24 hours to establish the total number of days, expressed as a decimal value, that the
vehicle was used. The median daily distance driven was 22 miles for the driving modes
survey. The comparable distance from the GM survey was 160 miles for a week, which
is about 22.9 miles per day.
Figure 6-16 and Table B-24 show the distribution of vehicle operation per
day. Vehicle operation represents the total time that the vehicle was operating on a
daily basis. For each vehicle, the data point used in this distribution is the average
vehicle operating time during a day. The median vehicle operation per day was about 60
minutes, which is equivalent to about 7 hours per week. In the GM survey, the weekly
vehicle operation was 7.1 hours.
6-30
-------
Figure 6-15. Daily Driven Distance Distribution
OS
I
if)
Q)
U
SZ
O
c
(D
O
>
o
ZJ
30 E
20 u
-------
Figure 6-16. Daily Operating Time Distribution
V)
0)
O
sz
c
(D
U
a>
Q_
19H
18
17-1
16-
15 -
14-
13-
12-
11 -
10
9
8
7-J
6
5-
4-
3
2-]
1
0-1
0
b 100
90
t m ' r 1 i 11' i1 i ' i 1 i ' i 1 i ' i 1 i ' i
i ' i ¦ i ¦ i ¦ i ¦ i ¦ i ¦ i ¦ i 1 i' i ' \ ¦ i ¦ i ¦ i ' i
1 234567891 1 1 1 1 t 1 i i
000000000012345678901 234567890
000000000000000000000
Vehicle Operation/Day (minutes)
-------
Figure 6-17 and Table B-25 show the distribution of the average number of
trips per day for the 216 vehicles. The median number of trips per day was about 6.2.
Figure 6-18, for major trips only, also shows a median of 6.2 trips per day, and the
median from the GM survey was about 3.9 trips per day. The plot has a fairly normal
distribution that relates well to the GM distribution.
Figure 6-19 and Table B-26 show the distribution of the average number of
stops per hour of vehicle operation for each of the 216 vehicles for all trips. The
definition of a stop is given in Section 6.1.2.1. Figure 6-19 shows that a near normal
distribution of the average number of stops per hour exists for the driving modes data.
The median number of stops per hour of vehicle use, when all trips are considered, is 38.
Considering that the median trip length is 7 minutes, this would mean there would be
about 9 trips per hour of use and, therefore, about 4 stops per trip. Figure 6-20 shows
the same distribution for major trips only.
Table 6-3F shows the average vehicle operation in minutes per day, vehicle
distance driven in miles per day, and the trips taken per day for the 216 vehicles. The
mean, standard deviation, minimum, and maximum of these measures are also presented.
These measures are weighted by the number of days that the vehicle was instrumented.
All the 3-parameter vehicles were instrumented for a full week, so if a vehicle was
instrumented on, for example, Tuesday, the datalogger on that vehicle was removed
Wednesday of the following week. Therefore, the actual number of vehicle days would
be close to eight days. For the 6-parameter vehicles, because of equipment constraints,
the datalogger installed on Tuesday was removed the following Tuesday; therefore, the
total number of days would be close to seven days. The 3-parameter vehicles, therefore,
have a slightly higher weighting in this table.
The weighted average vehicle operation time was found to be 68.5 minutes.
The average vehicle operation time from the GM survey was 72 minutes. The weighted
average distance driven per day was 28.1 miles, compared to 28.7 miles from the GM
6-33
-------
17
16
15
14
tn 1X
-------
Figure 6-18. Distribution of Number of Trips per Day for Major Trips Only
ON
I
Ui
18
17 H
16
15-
V) 14-
10
q
O y
^ 8
7H
6
b-\
4
3-
2-
1 -
0
c
O
' I ' I ' I
h 100
90
80
- 70
60
50
40
- 30
20
i ' i
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Number of Trips/Day
-------
Figure 6-19. Distribution of Number of Stops per Hour for All Trips
o\
t
u>
Q\
CO
0
O
*JC
d>
>
c
(D
O
a
5
30 £
20 u
10
0
Number of Stops/Hour
-------
Figure 6-20. Distribution of Number of Stops per Hour for Major Trips Only
12 H
11
10*3
tn
jD g.
• — ¦
_C
c
Q_
8
7
6
5
4
3
2
1
(H
n_n
n=
t11 ¦ i' i' i ¦ i' i' i ¦ \' i' i ¦ i ¦ i' i' i' i
i' i ¦ i • i • i ¦ i
i • i' i ¦ i' i' i' i
too
024681 1 1 1 1222223333344444555556
02468024680246802468024680
Number of Stops/Hour
-------
Figure 6-19. Distribution of Number of Stops per Hour for AH Trips
o\
I
u>
Ch
V)
o
sz
c
Q.
12-
11 -
10
9
8
7
6
5-1
4
3
2 -
1
0-1
i ' I ' I ' I ¦ I ¦ I ¦ i ' i ¦ i 1 i ' i ' i ¦ i ¦ i • i
i ' l ' I ' 1 ' i 1 i ' I ' i ' i ' i ¦ i ¦ l
» ' i • i » i
100
90
80
-------
Figure 6-20. Distribution of Number of Stops per Hour for Major Trips Only
12 4
f
Co
-J
if)
o
jc
a>
>
o
-t-j
c
a>
o
CD
Q_
h 100
i * i ' 1
I • i ¦ i ' I • i
24681 1 1 1 1222223333344444555556
02468024680246802460024680
Number of Stops/Hour
-------
survey. The average number of trips per day was 7.1, compared with 4.1 trips/day for
the GM survey.
Table 6-3G shows the average number of stops per hour of vehicle
operation for the 216 vehicles. This number has been weighted by the number of hours
of operation for each vehicle. Table 6-3H shows the weighted means and standard
deviations for the percent of accelerations, speeds, and specific powers greater than or
equal to the value shown. For example, the first line shows the average number of
vehicle operation observations in which the acceleration was ;>3 mph/s. On average for
all the vehicles, 3.26% of the vehicle operation observations had accelerations equal to
or above 3 mph/s. These measures have been weighted by the number of hours of
operation for each vehicle.
6.1.3 Comparison with Current FTP
An FTP data file was set up to simulate a vehicle experiencing a one-
second trip followed by a 12-hour soak, followed by a trip of 1369 seconds representing
the cold start and the hot stabilized bags of the FTP, a 10-minute soak, and then the first
505 seconds of the LA-4 representing the hot start bag of the FTP. This file was set up
just like the data file from any of the instrumented vehicles. Table 6-4 shows the speed,
acceleration, and specific power descriptions for the FTP. Table 6-5 shows the number
of seconds spent at any given speed and acceleration mode for the FTP. Note that the
column containing accelerations of 1 mph/s represents accelerations between 0 and 1.
Tables B-27, B-28, and B-29 show the distribution of speed, acceleration, and specific
power for the FTP. Figures 6-21, 6-22, and 6-23 show these distributions in graphical
form.
Figure 6-21 for the FTP can be compared with Figure 6-2 for the driving
pattern data. In general, the shape of the FTP speed curve is similar to that of the new
study, but the FTP speeds are shifted to lower speeds. The similarities include the
6-38
-------
Table 6-4
Speed, Acceleration, and Specific Power Statistics
for the FTP Cycle
Variable
Units
Number of
Observations
Mean
Standard
Deviation
Minimum
Maximum
MPH
mph
1874
21.2
15.94
0.00
56.70
ACCEL
mph/s
1872
0.00
1.41
-3.31
3.60
POWER
mph2/s
739
41.84
31.72
0.00
191.84
6-39
-------
Table 6-5
Speed/Acceleration Distribution for the FTP Cycle
SP5 AC1
Percent |
-3|
"2|
-1|
o|
H
2|
3|
4|
Total
o 1
0.05 |
0.32 |
0.53 |
18.06 |
0.00 1
0.00 I
0.00 I
0.00 I
18.96
5 1
1.34 |
0.27 |
0.59 |
0.43 |
0.27 |
0.91 |
0.59 |
0.43 |
4.81
10 |
1.34 |
0.43 |
0.21 |
0.48 |
0.11 |
0.32 |
0.37 |
1.55 |
4.81
15 |
1.01 |
0.96 |
0.32 |
0.32 |
0.43 |
0.43 |
0.75 |
0.96 |
5.18
20 |
1.01 |
0.91 |
0.85 |
1.55 |
1.44 |
1.50 |
0.75 |
0.96 |
8.97
25 |
0.48 |
0.91 |
1.39 |
5.29 |
5.34 |
2.46 1
0.91 |
0.21 |
16.99
30 |
0.32 |
0.32 |
0.43 |
6.78 |
6.41 |
1.44 |
0.11 |
0.11 |
15.92
35 |
0.00 |
0.11 |
0.75 |
3.26 |
2.99 |
0.59 |
0.21 |
0.11 |
8.01
40 |
0.00 |
0.00 |
0.32 |
2.35 |
1.50 |
0.21 |
0.11 |
0.00 |
4.49
45 |
0.00 |
0.00 |
0.53 |
0.00 |
0.00 |
0.32 |
0.00 |
0.00 |
0.85
50 |
0.00 |
0.00 |
0.11 |
1.71 |
0.96 |
0.11 |
0.00 |
0.00 |
2.88
55 |
0.00 |
0.00 |
0.21 |
2.78 |
2.46 |
0.11 |
0.00 |
0.00 |
5.56
60 |
0.00 |
0.00 |
0.00 |
1.50 |
1.07 |
0.00 |
0.00 |
0.00 |
2.56
Total
104
5.56
79
4.22
117
6.25
833
44.50
430
22.97
157
8.39
71
3.79
81
4.33
1872
100.00
6-40
-------
Figure 6-21. Speed Distribution for the FTP Cycle
100
90
80
70
- 60
50
40
30
20
10
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
Speed (mph)
-------
Figure 6-22. Acceleration Distribution for the FTP Cycle
50
CO
c
O 40
On
i
N)
o
>
25 30
-Q
O
20
O
-4—>
c
-------
Figure 6-23. Specific Power Distribution for the FTP Cycle
40
(/)
C
o
• I Ml
o
>
I—
(L>
in
o
*4—
o
-M
c
CD
o
ML.
I—I I 1 1 I I I I I—[¦
: 100
L 90
- 80
70
60
50
40
30
- 20
10
0
c
Q_
• -
O
D
E
15
o
50 100 150 200 250 300 350 400 450 500
Power (mph**2/s)
-------
similar percent of idling time and the general shape of peaks around 30 mph and around
55 mph. Some differences are the lower percent of speeds around 30-35 mph in the
driving pattern data, compared with the FTP. In addition, the higher speeds (>60 mph)
seen in the driving modes data are not included in the FTP.
Another interesting comparison is between Figures 6-1 and 6-24. Figure
6-24 shows an acceleration-speed contour plot for the FTP. One apparent difference is
the absence of accelerations above 4 mph/s and below -4 mph/s. In addition, the
general speed acceleration increase area (below 25 mph and 0-2 mph/s) seen in the
driving pattern data is not well defined in the FTP. Some of the high density areas
around zero acceleration and 30-35 mph and 50-55 mph are seen in both figures.
6.2 Data Biases
To assist in the evaluation of the driving pattern database, the methods of
data collection were examined for data biases. Three areas were considered: speed and
acceleration biases, vehicle selection biases, and any biases produced by the presence of
the datalogger on vehicles.
6.2.1 Speed and Acceleration Biases
The 3-parameter datalogger software used an algorithm to convert the
pulses received from the speed sensor into an average speed for each second that the
engine was on. As explained earlier, the datalogger measured the time between
consecutive pulses within a one-second window. The average of these times gave the
average time per pulse. During the calibration, the datalogger had stored the distance
per pulse, which was a characteristic of the speed sensor on that particular vehicle.
When this distance per pulse calibration factor was divided by the average time per pulse
in the one-second window, the average speed for that one-second window was obtained
and recorded in memory at the end of the one-second time period. To demonstrate that
6-44
-------
Figure 6-24. Acceleration/Speed Contours for the FTP Cycle
101
8-
I I I [ » » ' I | ' ' I I ) » » I I | I I ¦ I | I I I I | » » » < | f » ¦ f | ' I I I | I ' ' I | T » » » | I t I » | 1 I I I | I I I 1
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
Speed (mph)
-------
this technique provides an unbiased measure of vehicle speed, a specific numerical
example will be used.
Suppose the vehicle had a speed profile for a one-second period, as shown
in Figure 6-25. In this instance, the vehicle started from rest and accelerated to a speed
of 8 m/s in V2-second and then the speed was constant to the end of the second. The
acceleration for the first V2-second would be 16 m/s2 and the acceleration for the second
Vi-second would be 0 m/s2. Suppose that the speed sensor on this vehicle put out one
pulse for each meter that the vehicle traveled. The time when a pulse was generated
can be calculated from the relationship given below:
T = \J2 x s/a
where
s = the distance moved since time 0,
a = the acceleration, and
T = the time required to move distance s at acceleration a from time 0.
Using this expression, the times when pulses would occur, assuming that the first pulse
occurs at time 0, are shown in Figure 6-25. The pulses occur at 0, 0.354, 0.500, 0.625,
0.750, 0.875, and 1 second. The datalogger measures the time between these pulses,
which are 0.354, 0.146, 0.125, 0.125, 0.125, and 0.125 seconds, and determines the
average time per pulse. The average time per pulse is the sum of those differences,
which is 1 second, divided by the number of spaces between pulses, which is 6, to get
0.166 seconds per pulse. When the speed sensor calibration constant of 1 meter per
pulse is divided by the average time per pulse of 0.166 seconds per pulse, the result is a
calculated average speed of 6 m/s. This value of 6 m/s is the same value that an
observer of the speed profile in Figure 6-25 would say is the average speed of the vehicle
during the one-second time period.
6-46
-------
Figure 6-25. Speed Profile Example
1 m/pulse
0 -apulse
.146 .125 .125 .125 .125
» • • • t
0.5 1
Time (s)
-------
This example shows that the algorithm used by the datalogger to calculate
average speed for a second does not produce a bias. However, this average speed is
logged in the datalogger's memory at the end of the one-second time period, when the
instantaneous speed of the vehicle was not 6 m/s.
To calculate acceleration for data analysis purposes, the successive average
speeds for two seconds are subtracted, and the resulting acceleration value is assigned to
the last second. Bias in the calculated values for acceleration can be examined by
considering the fifth wheel data, which was discussed in Section 2. This comparison is
made between the accelerations calculated by the fifth wheel speed data, which were
collected and averaged every 1/7-second, and the 3-parameter datalogger speeds, which
were collected and averaged every second. Accelerations were calculated from both
dataloggers by taking the difference between successive speed values and dividing by the
time between the two readings.
The accelerations for Drive 3 for the three types of speed sensors are
shown in Figures 6-26, 6-27, and 6-28. In these figures, the solid line is the acceleration
calculated from the 3-parameter datalogger, and the dotted line is the acceleration
collected from the fifth wheel. The plots show that accelerations range from a high of
about 9 mph/s to a low of -14 mph/s. The driving for each of these drives was a hard
acceleration, followed by a panic stop. The three figures showed that there was very
good agreement between the acceleration calculated from the fifth wheel and the
datalogger. There is approximately a one-second shift in the 3-parameter acceleration
curves to times later than the fifth wheel acceleration curves, because of the backwards
differencing calculation used to determine acceleration for the 3-parameter logger data.
The figures show there may be a slight tendency for the 3-parameter datalogger to
underestimate the acceleration rates in comparison with those measured by the fifth
wheel.
6-48
-------
Figure 6-26. Evaluation of Calculated Accelerations for the Speedometer
Cable Speed Sensor
Time
-------
Figure 6-27. Evaluation of Calculated Accelerations for the Magnet Speed Sensor
Time
-------
Figure 6-28. Evaluation of Calculated Accelerations for the OEM Speed Sensor
20 H
Time
-------
Another feature noticed in these figures is that at accelerations near 0
mph/s, the fifth wheel accelerations fluctuate greatly up and down. In comparison, the
3-parameter datalogger does not exhibit these fluctuations. The point where accelera-
tions were 0 mph/s in the middle of the drive was when the vehicle was being driven at
approximately 50 mph. The driver was attempting to keep the vehicle going at a
constant speed at this point, but because the speed was high, there were significant
fluctuations in speed within each one-second time window of about 1 mph. When the
accelerations are calculated for the fifth wheel, these 1 mph speed fluctuations are
divided by the time between successive data points of 1/7-second, and this causes the
calculated value of acceleration to be very sensitive to the changes in speed of the
vehicle. During the wide-open throttle and the panic stop, the driver is not controlling
the speed closely. Vehicle speed is just rapidly changing in one direction.
Thus, this evaluation of the acceleration accuracy of the 3-parameter
datalogger indicates that the datalogger is accurate over a wide range of accelerations.
6.2.2 Vehicle Selection Bias
For the 3-parameter vehicles, a special vehicle selection technique was
developed and used in an attempt to get a representative sample of the vehicle popula-
tion that came to the I/M stations in Spokane and Baltimore. As discussed earlier, it
was desirable to sample the population in a representative fashion, so that different types
of vehicles and different types of drivers would be represented in the sample. If the
drivers of the vehicles did not want to participate in the project, or the vehicles could not
be instrumented for another reason, then these original vehicles were replaced with other
vehicles that were selected to have similar characteristics for classes of driver age,
vehicle age, origin, and type of vehicle.
6-52
-------
While 6-parameter vehicles were not selected randomly, it was desirable to
get a sample which, in some approximate way, matched the distribution of vehicle makes
in the eligible group of model year and make vehicles in the vehicle population.
Since the sampling of the population has been accomplished, it is now
possible to use statistical tests to determine if these samples are representative of the
population. Table 6-6 shows a list of the comparisons that will be made between pairs of
distributions for both cities and for the 3- and 6-parameter dataloggers as a function of
make, model year, and driver age class. This analysis was performed on SAS, using the
Chi2 Statistical Test for the categorical distributions of vehicle model, and using the
Mantel-Haenszel Chi2 Test for the ordinal distributions of model year. The results of
the statistical tests are shown in the last column of Table 6-6 and indicate there were no
differences between any of the distributions that were tested. Pairs of plots were made
to visualize the distributions that the statistical tests compared. These are discussed
below.
A tape for March 1992, giving the makes and model years of all vehicles
inspected in Baltimore at the stations where dataloggers were instrumented on vehicles,
was obtained from Envirotest, Inc. The tape was filtered to include only those vehicles
that had passed the emissions test or had been given waivers. This resulted in a
population of 17,279 vehicles. The distribution of these vehicles by make is shown in
Figure 6-29 and the distribution by model year is shown in Figure 6-30. The same
distributions were calculated for the vehicles that were instrumented in Baltimore with
3-parameter dataloggers. These distributions are shown in Figure 6-31 and in Figure
6-32. Comparison of the corresponding distributions for make and model year shows
that they appear to be very similar and, thus, support the result of the Chi2 statistic,
which indicates that there is no reason to believe there is a difference between the
distribution of the sampled vehicles and the distribution of the population. We have not
yet received a tape from V l'll, Inc. for the February 1992 population inspected in
Spokane, because VTTI is currently implementing major software revisions.
6-53
-------
Table 6-6
Results of Bias Analysis for Vehicle Selection
Site
Target
Vehicle
Parameter
Type
Distribution A
Distribution B
Parameter
Compared
Probability
Chi Square
Mantel-Haenszel
Baltimore
3
Instrumented Vehicles
Mar 92 Population
Model Year
-
0.75
Baltimore
3
Instrumented Vehicles
Mar 92 Population
Make
0.41
-
All Sites
3
Replacement Vehicles
Replaced Vehicles
Model Year
-
0.39
All Sites
3
Replacement Vehicles
Replaced Vehicles
Make
0.14
-
Baltimore
6
Instrumented Vehicles
Mar 92 Population of 89,
90, 91 Eligible Makes
Model Year
-
0.37
Make
0.26
-
On
1
L/i
4^
-------
Figure 6-29. Vehicle Make Distribution of Baltimore Eligible Population
PERCENT
17
16
15
14
13
12
11
10
9
7
6
5
4
3
2
1
0
AAABBCCCDEFGMH I JLMMMMNOPPPRSSTVV
CMUMUAMHOAOUOYSA I A E E I I LLOOEAUOOO
UEDWI DER0CRCNUJGN2RRTSDYNRNABVI L
R R I CIVYGLD DN2UCOC2SSSMTSABAOKV
Make
-------
Figure 6-30. Model Year Distribution of Baltimore Eligible Population
PERCENT
1*77 197S ttTt 1MO 1Mt IMS 1**3 IM4 IMS 19M 1M7 1M* 1M* 1M0 INI IM1
Model Year
-------
Figure 6-31. Vehicle Make Distribution of 3-Parameter Baltimore Instrumented Vehicles
Make
-------
Figure 6-32. Model Year Distribution of 3-Parameter Baltimore Instrumented Vehicles
Model Year
-------
Another comparison that is important to make is that of the originally
solicited vehicles, which for some reason were not able to be instrumented, with the
distribution of vehicles that replaced them. There were 120 3-parameter vehicles that
were solicited and could not be instrumented for one of a number of reasons, and there
were 90 3-parameter vehicles that were replacements for a portion of these vehicles.
Figures 6-33 and 6-34 show the distributions of vehicles that were not able to be instru-
mented by make and model year. Figures 6-35 and 6-36 show the same distributions for
the vehicles that were finally instrumented for vehicles that could not be. Visual
comparison of these distributions tends to confirm the results of the statistical tests that
the distributions were not significantly different.
The 6-parameter dataloggers were placed on vehicles of the seven
manufacturers for model years 1989, 1990, and 1991. The inspection/maintenance
record tape for March of 1992 in Baltimore was filtered to determine the eligible makes
and model years of vehicles that met these criteria. The make and model year
distributions of these Baltimore vehicles which were eligible for 6-parameter installations
are shown in Figures 6-37 and 6-38. The make and model year distributions for the 6-
parameter vehicles that were instrumented at Baltimore are shown for comparison in
Figures 6-39 and 6-40. Comparison of the population and sample distributions for make
and model year using the Chi2 statistic and the Mantel-Haenszel statistic indicate that
there is no significant difference between these distributions. However, the distribution
comparisons for make do not appear to be all that similar. The sample would have been
expected to have more Toyotas, more Fords, and fewer Dodges than it had.
6.2.3 Datalogger Presence Bias
One of the disadvantages of performing a private vehicle instrumentation
study is that there is a question of whether the driver of the vehicle drives differently
because he knows he is being monitored. This is not a problem for a chase car study.
To try to get an estimate of whether people drove differently because they knew
6-59
-------
Figure 6-33. Vehicle Make Distribution of Solicited 3-Parameter Vehicles
That Were Replaced
PERCENT
Make
-------
Figure 6-34. Model Year Distribution of Solicited 3-Parameter Vehicles
That Were Replaced
PERCENT
1*77 lt7» 117* 1MO 1M1 1M2 1M3 1884 IBM 1966 1M7 1986 1MB 1»M IM1 1M2
Model Year
-------
Figure 6-35. Vehicle Make Distribution of 3-Parameter Replacement Vehicles
PERCENT
AAABSCCCDCFGHH I JJLLMWMMNOPPPftSSSTVV
CMUMUAHHOAOMOYSACC IACC I I LLOOEAUUOOO
UEOWIOCROCRCNUUCCXNZRRTSOYNRNABZYLL
* « I CIVYGLD ON2UPUCDCZSSSUTSABAUOKV
Make
-------
Figure 6-36. Model Year Distribution of 3-Parameter Replacement Vehicles
Model Year
-------
Figure 6-37. Make Distribution of 6-Parameter Eligible Baltimore Population
PERCENT
BUtC CMCV cum OOOG F0*0 UNC MUD MCTC WTS WSS OLDS P|.*WI PONT TOTO
Make
-------
Figure 6-38. Model Year Distribution of 6-Parameter Eligible Baltimore Population
ON
ON
PERCENT
60
••77 1978 1M0 1961 1982 IMS 1904 IMS 1966 1967 1966 1999 1990 1991 199?
Model Year
-------
Figure 6-39. Make Distribution of 6-Parameter Baltimore Instrumented Vehicles
PERCENT
30-1
euc CHCV CHfty DOOC row) UNC MAZO WOK HITS NftS OLDS PLYM PONT TOYO
Make
-------
Figure 6-40. Model Year Distribution of 6-Parameter Baltimore Instrumented Vehicles
Model Year
-------
dataloggers were on the car, the data was analyzed by observation phase during the
week.
The idea here is that it is possible that, after a certain length of time, the
driver of the vehicle may forget that the datalogger is on the vehicle and, therefore,
would drive in a normal fashion. The observation phases were picked such that two
evenings passed before the second observation phase began. The observation phase
parameter had two levels. The first phase occurred between the time the datalogger was
installed on the vehicle and late in the night of the first full day of driving with the
datalogger on the vehicle. The second observation phase occurred from the end of the
first observation phase to the end of the data acquisition period for that vehicle.
The driving patterns database was analyzed by observation phase for all
trips (minor and major). The results of this analysis are included in Volumes 2 and 3.
For the purposes of discussion, a few of the results are shown here in the body of the
report. Three pairs of plots are shown to compare the effect of observation phase on the
acceleration/speed contour, the trip duration, and the soak duration.
The acceleration vs. speed profiles for the two observation phases are
shown in Figures 6-41 and 6-42. These contours appear to be very similar. The
maximum and minimum accelerations occur at the same speeds and have approximately
the same values. The shadings on the different portions of the contours occur in
approximately the same place. One noticeable difference, which may or may not be
significant, is that in the second observation phase, there may be more observations at
speeds higher than 65 mph.
A comparison of the trip times for the two observation phases can be made
by looking at Figures 6-43 and 6-44. The distribution in the second observation phase
appears to be smoother. This is probably because there are more observations in the
second observation phase. However, the general shape of the distribution-the height of
6-68
-------
Figure 6-41. Acceleration/Speed Contour for First Observation Phase
lOn
8-
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
Speed (mph)
-------
Figure 6-42. Acceleration/Speed Contour for Second Observation Phase
Speed (mph)
-------
Figure 6-43. Trip Time Distribution for First Observation Phase
On
I
-J
10-
9-
8-
V)
Cl
7-
v_
1—
6-
O
b -
c
3-
CL
2-
1
0-
0.2
» » i « > t i
¦ i i i i i 11
2.0 20.0
Trip Time (minutes)
-100
- 90
• *' r
200.0
-------
Figure 6-44. Trip Time Distribution for Second Observation Phase
OS
K>
12-
11 •
10-
(/)
9-
Q_
V_
1
8-
1
v»_
7-
o
4-1
6
r
(D
b
O
4-
a>
Q_
3-
2-
1 ¦
0-
0.2
100
90
80
70
I f T I < I I |
I * i » f y
c
a>
o
o
ZJ
E
20 O
10
0
2.0 20.0
Trip Time (minutes)
200.0
-------
the peak, the location of the peak, and the width of the peak-are similar for both
phases. Thus, there is no apparent difference in the lengths of trips between the two
phases. This might be expected, since it may be unlikely that people would change the
trips they take just because a datalogger is on their vehicle.
A comparison of the soak times before trips is made for the two observa-
tion phases in Figures 6-45 and 6-46. There are some differences between these
distributions. The most obvious one is that the overnight soak periods which have a
peak around 800 minutes are less prevalent compared to the non-overnight soak times
(between 3 and 100 minutes) than in the second observation phase. There also appears
to be a large number of short soak times in the second observation phase around 0.3
minutes, which is not present at all in the first observation phase.
Additional comparisons can be made for the first observation phase vs. the
second observation phase by considering the other plots and the calculated statistics
which are shown in Volumes 2 and 3. Based on a visual examination of these, it does
not appear that there are any major reasons to believe that driving during the first
observation phase was different than driving during the second observation phase.
6.3 Options for the Advanced Analysis and Simulation of Driving Patterns
Several authors have published studies of the data analysis of light-duty
vehicle driving patterns (6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16). Two options are presented
below for data analysis and for the simulation or selection of driving cycles that could be
used in a new Federal Test Procedure. The first option discusses the use of advanced
statistical techniques for analyzing and characterizing both the details of trips and the
soak behavior of vehicles in the driving pattern database. The final result of this
advanced statistical analysis would be the selection of existing driving segments in the
database that would represent the driving of the vehicles in the database as a whole.
Because these selected driving segments would have all of the fine structure that real
6-73
-------
Figure 6-45. Soak Time Distribution for First Observation Phase
8 -I
OS
4*.
If)
7-
"O
o
*L_
6-
a>
Q_
5-
o
o
LO
4-
o
3-
c
a>
2-
o
a)
i -
Q_
0-
T~
4
0.1
I • « f « * « (|
1.0
* V 1 I V I I VI
I I t I I V I I
I T I I I I I f I
I i i 1 r i l I
100
90
80
70
60
50
40
30
20
10
0
c
Q_
Q>
>
O
3
E
3
u
10.0 100.0 1000.0
Soak Time (minutes)
10000.0
-------
Figure 6-46. Soak Time Distribution for Second Observation Phase
OS
t
-J
8H
O
6
-------
vehicles display, an informed selection of the segments has a good chance of being
representative of the driving behavior of all the vehicles as a whole. However, it is not
at all obvious that a single driving segment for a particular vehicle is the best way to
simulate overall driving. There may be great advantages in using two or more driving
cycles in a new Federal Test Procedure.
Another approach to developing driving cycles is also presented in the
following paragraphs. This approach uses a combination of signal processing analysis
followed by neural network simulation of the soak and detailed speed behavior during
trips of all of the data in the database to arrive at a system for simulating driving cycles
that are representative of overall driving behavior. The simulations produced by such a
system would have all of the features exhibited by the database as a whole; however,
none of the simulations would be exactly like a driving segment contained in the
database.
The test of whether driving segments selected from the database using
advanced statistical techniques or simulated driving cycles produced by a signal
processing/neural network system are representative of the driving patterns in the
database must be made by using any and all statistics to compare the resulting candidate
cycles with the driving pattern database.
6.3.1 Advanced Statistical Analysis
The summary statistics calculated earlier in the section for the driving
patterns database and for the current FTP cycle can be used to evaluate candidate
driving cycles. However, to get a better understanding of driving behavior and, perhaps
more specifically, the long time period driving behavior, other, more advanced statistical
techniques must be used. Since these techniques lead to better understanding, they
should lead to a more insightful basis for constructing or choosing candidate driving
cycles that are representative of the driving pattern behavior in the database.
6-76
-------
The discussion below provides two advanced statistical techniques that
could be used to select trips and soaks which could be demonstrated to be representative
of observed driving patterns.
6.3.1.1 Identifying Types of Trips
To determine the trip or trips to be used for a new certification test
procedure, it is desirable as a first step to determine the distinct types of trips that occur
with reasonable frequency. The method of addressing this issue will be to examine the
trips for which speed vs. time data are available in the database. The objective will be
to divide these trips into subsets, or clusters, so that the trips in each cluster are basically
similar but differ somewhat in detail.
Radian's preliminary analysis of the data from EPA's Operational Charac-
teristics Study (17) suggests that there are at least two types of trips. The two types of
trips identified in that study had the following characteristics:
• Type A. In each trip in this set, there was at least one period of 100
seconds or more with speeds near 65 mph without a sustained
acceleration or deceleration. A "sustained acceleration" was defined
as an interval with monotonically increasing speed with a total
change in speed of at least 20 mph. A "sustained deceleration" was
defined analogously. In this set of trips, there were also periods of
rapid alternation between acceleration events and deceleration
events.
• Type B. In each of this set of trips, there was an absence of either
very long or very short intervals between acceleration events.
The general features of the speed vs. time curves were visually similar
within either type and were clearly different between the two types. It was shown that
either the average or the standard deviation of the time between acceleration events
would have been an adequate measure by which trips could have been divided into two
6-77
-------
groups. These two measures were used for illustration only and were not intended to
represent a full characterization of the trips. The FTP currently used for emissions tests
is essentially of Type B (17).
While a visual examination of plots of speed vs. time is very important, it is
not feasible to examine every trip in the database from the current study in this way. In
the following section, quantitative measures that can be used as a basis for dividing trips
into "clusters" or "types" are discussed. Some of these sets of measures involve large
numbers of variables. Following that section, a method (principal component analysis)
for reducing the dimensionality of the mathematical space that must be handled without
significant loss of information is discussed. Next, the grouping of the trips into clusters
with similar features is discussed. This operation is called cluster analysis.
Both principal component analysis and cluster analysis have been widely
used, and a body of literature exists for each. The major emphasis in this section, then,
will be placed on selecting a meaningful set of measures that can be used to characterize
the automobile trips.
Characterizing Measures for the Trips-There are many quantitative sets of
measures that could be used to characterize vehicle trips. We will discuss two general
classes of sets. First are measures defined on the basis of engineering judgement. It
would be ideal to define a small set of measures of this type that contained all the
essential information about the trips. The danger, however, is that an essential feature
of the trips might not be quantified by the engineering measures. Second are larger sets
of measures intended to avoid this difficulty. These types of sets will be called
"nonprejudicial," since there is probably a greatly reduced chance that a priori judgement
will limit the information contained in them.
In any case, selection of the characterizing measures should not be strictly
based on the emission properties of existing automobiles. Changes in the design of
6-78
-------
future cars may alter the relationship between speed and acceleration, and emissions.
The goal is to develop measures that characterize the speed vs. time profiles in a general
way, but not necessarily in a way closely related to the emission properties of current
automobiles.
Sets of Trip Measures Based on Engineering Judgement-There are many
possible sets of measures. As discussed earlier, either the mean or standard deviation of
the time between acceleration events would have been adequate to distinguish between
trips of Types A and B, defined in Section 6.3.1.1. These measures, however, were
intended only to illustrate the fact that trips can be quantitatively clustered in meaningful
ways.
Maurin and Crauser (6) characterized French driving patterns with the
following set of measures:
• The trip duration;
• The distance covered;
• The average speed;
• The maximum speed;
• The number of extrema, relative to speed;
• The idling time;
• The percentage of the time the speed was relatively stable
(acceleration within specified limits);
• The percentage of the time the speed was between 0 and 15 km/h;
• The percentage of the time the speed was between 15 and 45 km/h;
6-79
-------
• The percentage of the time the speed was between 45 and 75 km/h;
and
• The percentage of the time the speed was greater than 75 km/h.
While the measures above seem to be reasonable, it would be interesting
to include a more detailed characterization of the accelerations and possibly of the
combinations of speeds and accelerations that occurred during the trip. It is true,
however, that the number of extrema in speed implies the number of separate
acceleration or deceleration events. These observations illustrate the fact that any set of
measures based on engineering judgement may fail to reflect some important characteris-
tic of the trips, which leads to the next topic, nonprejudicial sets of measures.
Nonprejudicial Sets of Measures-One set of nonprejudicial measures
characterizing the trips would be the ordinates of the bivariate distribution of speed and
acceleration. An example of this type of bivariate distribution is presented in Figure 6-1.
While engineering judgement was not used in detail to specify every measure in the set,
this set of measures is consistent with engineering judgement, since it is reasonable to
expect that combinations of speed and acceleration will affect emission rates.
The bivariate distribution characterizes the frequencies of occurrence of
different conditions but does not explicitly characterize the time ordering of events. This
information is, at least to some extent, inherent in the distribution, however. Consider,
for example, a speed maintained for one or more extended intervals during the trip.
This can only occur if the bivariate distribution includes a relatively high frequency of
occurrence of this speed concurrent with accelerations that are small in absolute value.
Notice that the bivariate distribution shown has 35 levels in speed and 50
levels in acceleration, resulting in 1750 cells. Thus, the set of measures consists of 1750
variables. Ways of reducing the dimensionality without significant loss of information are
discussed in the following subsection.
6-80
-------
Frequency-domain techniques offer a different approach. In this context,
frequency-domain methods are those based on the Fourier transform, and the word
"frequency" should not be confused with "frequency of occurrence" of events in the
bivariate distribution discussed above. Use of the power spectrum is one approach that
has an intuitive appeal. There is a good chance that the power spectrum would provide
a meaningful characterization of the trips. It would be possible to include: 1) the power
spectrum and phase, or 2) both the real and imaginary components of the Fourier
transform in the set of measures. If frequency-domain methods are employed, we
recommend that the spectrum be used as a first step to determine whether meaningful
trip clusters can be obtained on this basis. If not, an alternative approach could be
devised. The mathematical intractability of phase for our purposes is circumvented,
however, by the use of the bivariate distribution instead of frequency-domain methods to
determine a set of characterizing measures.
Methods for Reduction of Dimensionality—Regardless of the method used
to characterize the trips, it is likely that the number of variables selected will be larger
than the minimum needed to represent the essential information. Even a set of
measures based on engineering judgement may contain some redundancy.
Consider, for example, the bivariate histogram shown in Figure 6-1. Notice
that, to a large extent, adjacent cells tend to be similar; i.e., usually both are higher than
average or both are lower than average. That is, the frequencies of occurrence
represented in adjacent cells do not contain independent information. Moreover,
certain cells will always or almost always have zero frequencies. Cells with the lowest
speed category and with negative accelerations with large magnitudes provide one
example. Cells with the maximum speed category with accelerations with the largest
absolute values represented are another example.
Thus, it is evident that the number of truly independent dimensions of
information is less than 1750, the number of variables associated with the bivariate
6-81
-------
distribution shown in the figure; the ordinate corresponding to each cell is a separate
variable. One way to address this issue is to eliminate certain cells altogether and to
combine other groups of adjacent cells.
Principal component analysis is a more general approach that would apply
in the case of any of the methods discussed above for characterizing the trips. Principal
component analysis is often used when the variables are not statistically independent;
therefore, groups of two or more variables, to some extent, contain redundant
information. Principal component analysis produces a smaller, derived set of variables
that are statistically independent and that contain the essential information represented
in the original set of variables.
The first step in principal component analysis is to determine the linear
combination of the original variables that has the largest variance (for each principal
component, there is a normalization constraint on the coefficients ~ otherwise the
variance of a component would be unbounded). This linear combination is a new
variable and is the first principal component. The second step is to determine the linear
combination of the original variables that has the largest variance of any combination
statistically independent of the first principal component. The second linear combination
is the second principal component. Successive linear combinations are determined, each
statistically independent of the preceding ones.
There are as many principal components as original variables. Suppose,
however, there are 1750 original variables, but that virtually all of the essential
information can be expressed in ten appropriately selected independent dimensions.
One would expect, then, that the first ten principal components would adequately
represent the original set of variables.
The variances of the principal components can be used to determine the
number of components that must be used to retain, for example, 95% of the original
6-82
-------
variance in the system. The original variance in the system equals the sum of the varian-
ces of the original variables, which also equals the sum of the variances of the complete
set of principal components. It may seem that the sum of the variances of the original
variables would not be meaningful if those variables had different units (e.g., speed and
acceleration). This problem is resolved by transforming each variable to a pure number
with zero mean and unit variance. Subsequent analysis, then, can be performed using
the retained principal components in place of the original variables.
Earlier in this section, a set of 11 measures used in a study reported in the
literature was presented. While it is obvious that the 1750 variables associated with the
bivariate distribution contain redundancies, it is less obvious that this set of 11 variables
contains significant redundancies. Nevertheless, an analysis revealed that 80% of the
variance in the system can be retained by employing the first four principal components.
Cluster Analysis-Cluster analysis is used to divide a set of objects into
subsets, or clusters, so that the objects are similar within the clusters and so that there is
a meaningful difference between any two of the clusters. Here, each object is one of the
trips in the database. Each cluster, then, will correspond to an identifiable type of trip,
such as Type A or Type B, discussed above.
Cluster analysis requires a value or vector of values, which characterizes
the different trips. As discussed above, there are many possible choices for this vector.
One example is the set of retained principal components derived from the bivariate
distribution of speed and acceleration. More generally, the vector would consist of the
subset of the principal components required to retain the essential information in
whatever set of original variables had been used.
To divide the trips into clusters, it is necessary to define a quantitative
measure of the degree of dissimilarity between any two trips. This measure is necessary
to determine whether two trips are similar enough to belong in the same cluster or
6-83
-------
whether they should be placed in different clusters. The Euclidean distance between two
vectors of principal components is one possible measure of dissimilarity. Cluster analysis
is widely used, and documentation is readily available (18). For this reason, the
computational details of the different types of cluster analysis will not be discussed here.
6.3.1.2 Selection of Trips for Driving Cycles
It is desirable to be able to select one or more trips to be used in a given
emissions test. Two approaches will be considered. First, randomly selecting trips from
the database is discussed. Second, generating trips by modeling is discussed.
Selection of Trips from the Database-Suppose that a trip corresponding to
a specific one of the clusters is required for an emissions test. Suppose further that there
are N trips in the database corresponding to this cluster. One approach, then, would be
to randomly select one of these trips to be used in the test.
This approach would apparently satisfy the objectives stated in Section
6.3.1.1 reasonably well. We assume that the trips in the cluster are similar in general
features but differ in specific details. For example, the maximum acceleration, idling
periods, and periods of steady speed would not occur at the same times within the
different trips within a cluster.
This approach is simple, and it has certain advantages over modeling. The
trips represent actual trips; there is no concern about whether some feature has not been
retained through the modeling process. The shape of the speed vs. time curves during
sustained accelerations, for example, will retain the features and the variability within
trips and among trips actually observed in the field.
6-84
-------
On the other hand, the number of trips for a given cluster is limited to
those in the database, unless further testing is performed to add to the database.
Modeling, discussed in the next subsection, provides a way to address this issue.
Generation of Trips by Modeling—Monte Carlo simulation has the
advantage of being able to generate unlimited numbers of trips that vary in specific
detail but that have similar general features. It is very important, however, to test the
method thoroughly to ensure that the trips it generated possess the desired general
features. This can be achieved in part by visually comparing speed vs. time plots for sets
of actual and modeled trips. The comparison of various quantitative measures is also
important. Statistical hypothesis tests can be performed to determine whether the actual
and modeled trips differ significantly with respect to average speed, average acceleration,
and even the statistical distribution of speed or acceleration.
After a trip has been randomly generated for an emissions test or other
use, it is desirable to test the trip to ensure that it is not a probabilistic anomaly that is
not representative of the type of trips of interest. To give an extreme example, many
types of Monte Carlo models allow a very small but non-zero probability that the entire
trip is spent idling. It would be possible to compute a number of quantitative measures
(e.g., average and standard deviation of speed and acceleration) to determine whether
they fall within a reasonable set of limits determined from the trips in the cluster of
interest in the database.
Two general approaches for modeling are discussed here. In the next
section, an approach is discussed that is based on identifying specific types of features, or
events, that occur within a trip. Examples of these events include idling, sustained
acceleration, sustained deceleration, and traveling at a relatively steady speed. These
events can be called "lumped features," since each represents the combined speed vs.
time data for a period of time during which a specific type of activity is taking place. In
6-85
-------
the following section, an approach is discussed that does not require the identification of
lumped features. Other approaches are briefly discussed in the last section.
Modeling Based on Lumped Features—The first step in the approach based
on lumped features is to identify the features to be used. Second, it is necessary to
identify occurrences of these features in the trips from which the model is to be derived.
Because of the large number of trips in the current data base, this would probably
require the development of software that searched each trip to identify the features
present and their times of occurrence.
The next step is to develop a model that generates trips that have the
general features of those in the database. One approach is to use a Markovian transition
model. In this type of model, a series of states (or lumped features) is generated in their
order of occurrence; the complete set of events, then, defines the trip. Given the set of
transition probabilities for all states that can possibly follow the current state, the next
state can be randomly generated. If the current state is idling, for example, the possible
following states might include sustained acceleration and low and variable speed.
Features of certain states have to be characterized from the database. The
generation of an acceleration event requires the duration of the event, the final speed,
and the shape of the speed vs. time curve during the acceleration.
A Monte Carlo model of the type discussed above was developed by Smith
(7), who found that all decelerations followed the same general curve. He identified
three different acceleration curves, however, depending on the total change in speed and
depending on whether the automobile was initially at rest or not. Smith's study did not
accomplish all the goals proposed here, however, since clustering to identify different
types of trips was not part of his study.
6-86
-------
Modeling Without Identifying Lumped Features—An alternative to the
approach above is to generate a speed vs. time history by a Monte Carlo approach
without identifying the lumped parameters. An advantage to this approach is that a
greater variety in the small details of the features results. The shapes of the acceleration
curves, for example, would not be limited to the finite number of shapes selected for
inclusion in the model in the lumped features approach.
In the approach discussed in this section, the speed for the next time point
is randomly generated on the basis of the speeds at the current and past times. First, it
is clear that an autoregressive moving average (ARMA) model would not work. An
ARMA model has the following form:
m
= ^ \ Si-k + eitl
k=0
where
Sj = speed at the ith time point,
ak = kth coefficient in the model, and
e; = random component introduced at the ith time point.
This model generates a stationary time series, i.e., one whose statistics do
not change with time. This model would not be expected to produce a nonstationary
time series with distinct idling periods, acceleration periods, periods with relatively
constant speed, etc.
A transition model is a better choice. If the current speed is zero, for
example, a realistic probability that the speed would again be zero at the next time point
would be used. Unlike the ARMA model, this scheme would produce idling periods. It
is clear, however, that a Markovian model would be inadequate. A transition model is
Markovian if the next state depends only on the current state and the transition
probabilities, and not on past states. That is, a Markovian model is memoryless.
6-87
-------
To see why a Markovian model would not work, consider the case in which
the current speed is some intermediate value, e.g., 18 mph, which is not normally
sustained for any period of time. Suppose this state is reached by an acceleration from
0 mph. If a Markovian model is used, there is nothing to tell the model that an
acceleration is taking place; therefore, the next speed should be greater than the current
speed. Thus, this type of model would not be expected to produce sustained ac-
celerations or sustained decelerations.
Suppose the model randomly generates the next speed, basing it on the
current speed and one or more previous speeds. The two or more resulting speeds on
which the transition is based, then, would include information about the direction of
change of speed. If speeds i-2, i-1, and i are 20, 22, and 24 mph, respectively, then the
probability that the next speed will be 26 mph is high. Slight variations in the rate of
acceleration are possible, however, assuming these variations exist in the actual trips
from which the model is derived.
A disadvantage of this approach is that, in order to build a transition
matrix, you must discretize the speeds. As the number of speed categories increases, in
one sense, the effects of discretization are mitigated, which is desirable. Increasing the
number of categories also increases the size of the transition matrix and decreases the
sample size in each discrete category. Thus, several considerations must be balanced
when determining the size of the speed categories, or bins.
Increasing the number of past speeds on which the transitions are based,
increases, to a point, the validity of the model. Past some point, one would expect
diminishing returns. Increasing the number of past speeds used would also increase the
size of the transition matrix.
The size of the transition matrix could easily get out of hand. Suppose
speed was divided into 30 bins and that the next speed was a function of the current and
6-88
-------
past speeds. The total number of entries in the transition matrix, then, would be 303, or
27,000. It is clear, however, that most of the transitions would be nonsensical. If the
current and past speeds were both 0 mph, for example, the possibility that the next speed
would be 60 mph could be ruled out, assuming a reasonably small time step, such as one
second. Perhaps the transition matrix could be designed so that, for each current and
past speed, only the most likely m following states were represented. If there were 30
speed bins, and if m were 6, then this mechanism would reduce the size of the transition
matrix from 27,000 to 5,400 entries.
Other Approaches-The above discussion is intended only to provide a few
of the multitude of possible ideas about Monte Carlo modeling. Many approaches that
are entirely different from those discussed above are possible, as are combinations of the
different approaches discussed to create a hybrid approach.
One example is to combine the lumped features model with ARMA
techniques. The transition model based on lumped events could be used to identify the
sequence of broadly defined events, their durations, and certain other characteristics.
Within a period of relatively steady speed, however, an ARMA model could be used to
model the small variations in speed that would be expected.
If the Monte Carlo modeling is undertaken, further ideas will be
considered before the model is completed. Inevitably, refinements of the model will be
achieved during the testing process. It is important that the final model be sufficiently
detailed to produce realistic trips. Unreasonable complexity beyond that needed to
achieve this goal, however, should be avoided.
6.3.1.3 Characterizing Soaks
The term "soak" will be used to refer to a period during which a car is not
in use. The duration of a soak equals the time the engine has to cool after its recent
6-89
-------
use. The temperature of the engine when it is next turned on affects the emissions
during startup of today's vehicles. It is well known that start-up emissions are greater
during a cold start. The soak duration, therefore, is an important factor when designing
a new FTP.
Characterizing Soak Durations-The database contains the results of tests
of 216 automobiles lasting one week each. Twelve thousand soaks occurred during these
tests; therefore, the sample size is large and it should be possible to characterize the
statistical distribution of the durations very well (See Figure 6-7).
It is important to characterize the distribution of short soak durations
accurately. It is known that the rate of change of the temperature of an object is a
monotonic function of the difference between its temperature and the temperature of its
surroundings. Thus, the rate of change of the temperature of the car's engine is greatest
just after it is turned off. As the engine continues to cool, the rate of change of
temperature decreases (in absolute value). Thus, there might be an important difference
between a 15-minute and a 30-minute soak. The difference between a three-hour
duration and a three-hour-and-15-minute duration is not as important, although, as
before, there is a 15-minute difference between the two durations.
Differences between soak durations beyond 12 hours are usually considered
unimportant, because the vehicle's engine should have essentially reached the ambient
temperature for any reasonable set of circumstances (ambient conditions, etc.). The
database contains for 6-parameter data the coolant temperature up until the time the
engine was turned off. The temperature when the engine was just turned on is also
known. These temperatures can be used to determine the coolant temperature rate of
cooling after engine shutdown. This will be useful in determining the soak time
categories for modeling.
6-90
-------
Given these preliminary comments, the statistical distribution of the soak
durations is believed to be of primary importance. There are various ways to
characterize this distribution. One representation is based directly on the set of observed
soak durations. First, the durations must be ordered:
ti * t2 s . . . <; tn
where n = 12,000, the number of soak durations (tj) in the database. Then the correct
way to construct the empirical cumulative distribution on the basis of the complete set of
points is as follows:
P(t <; t^ = i/(n+1)
where P() denotes the probability of occurrence of the indicated event.
Notice that, by this method, we correctly avoid claiming to have estimated
the absolute maximum or the absolute minimum of the distribution on the basis of a
finite sample. The maximum, for example, is a value tmax such that P(t ^ tmax) = 1. In
the approach above, the probability associated with the largest value in the database is
n/(n+1). Since this probability is less than one, we have not represented tn as the
absolute maximum of the distribution. Actually, there is no apparent reason to say there
is an absolute maximum soak duration. The statistical details of this method for
estimating an empirical cumulative distribution function are discussed by Gumbel (19).
This approach is appealing in that no information from the original data
set is discarded. It is unappealing in that storing the cumulative distribution requires a
large amount of information. However, the information can be compressed to some
extent, since the maximum soak duration of practical interest is known to be 12 hours.
6-91
-------
Alternatively, it would be possible to define a set of bins and count the
number of soak durations in each bin. A probability distribution could then be
determined on the basis of these counts. As discussed above, small differences between
durations become decreasingly important as the duration increases. On the basis of this
consideration alone, it might be advantageous to use a small bin size for small soak
durations and a larger bin size for larger durations. For example, bins based on a log
time scale, as shown in Figure 6-7, might serve this purpose.
The specific features of the data set should influence the selection of bins,
however. It has been observed that the distribution of soak durations has a broad
plateau from about 3 minutes to 100 minutes and a second peak near 700 minutes. The
latter corresponds to overnight soaks. It might be advantageous to select one or more
smaller bins near 700 minutes, for example, to characterize the peak near that point.
Before the actual set of bins is determined, it should be verified that the
bins are sufficiently narrow to represent the essential information from an engineering
point of view. The bins should not be so narrow, however, that there are very few data
points in each bin; given the large database with 12,073 soak durations, however, this
does not appear to be a limiting factor.
Suppose, then, that n; data points fall in Bin i. The probability of
occurrence of a soak duration within the limits of Bin i is n^n, where n= 12,073.
Calculation of the probabilities for all bins, then, would produces a description of the
distribution of soak durations.
This representation of the soak durations is very appealing in that the
amount of information that must be stored is much smaller than the original database.
On the other hand, some information is lost when data are binned. If the bins are
sufficiently narrow for practical purposes, however, the loss of information should not be
important.
6-92
-------
Another possible way to represent the distribution, which is mentioned
here for completeness, is as a theoretical type of distribution (normal, log normal, etc.).
A log-normally distributed variable is nonnegative and has no upper bound; soak
durations have both of these properties. The log normal distribution is positively
skewed; that is, the random variable has a predominance of moderate values, occasional
values much larger than the mean, and an absence of corresponding values much less
than the mean. The first step in determining whether such a distribution is appropriate
is to plot a histogram of the soak durations and determine whether it has the appearance
of the standard distribution. If so, a formal hypothesis test (e.g., a Kolmogorov-Smirnov
test) is performed to determine whether the empirical distribution matches the
theoretical distribution within random variability and within the practical requirements.
As mentioned above, however, the histogram of soak durations has a broad
plateau between 2 minutes and 200 minutes and a peak near 700 minutes. In view of
these features, we believe that the use of a theoretical type of distribution will probably
not be feasible.
Other Considerations-Another consideration is the time of day the soak
duration occurs. Whether a soak occurs during daytime or nighttime has an important
effect on evaporative emissions but not on emissions from the tailpipe. The concern
here pertains to tailpipe emissions.
Another consideration pertains to the type of trip that follows soaks of
different durations. A drive to work would typically (but not always) follow a long,
overnight soak, for example. However, a drive to work could be a seven-minute drive
from one's home to a nearby office or a 45-minute drive across a major city. It seems
that almost any type of trip could follow a soak of any duration.
6-93
-------
6.3.1.4 Selection of Soak Durations to be Used in a Given Emissions Test
The characterization of soak durations described in the preceding section
will help identify what needs to be included in an emissions test regarding durations.
Three basic approaches for determining the durations to be used are 1) select a set of
two or more durations to be used in all tests, 2) randomly generate the durations to be
used in each test, and 3) use a combination of 1) and 2).
If the same set of soaks is to be used in all tests, the set can be determined
partly on the basis of engineering judgement and partly on the basis of the statistical
distribution. Since there is clearly an interest in cold starts, the first soak could be one
long enough to allow the engine to reach ambient temperature. If only one other
duration is used, the median determined from the database is a reasonable choice. If
three others are used, the quartiles (25th, 50th, and 75th percentile values) are reasonable
choices. If the number of soaks to be used is more than two, the soaks to be used after
the first, long soak can be applied in random order.
In the previous section, a number of possible ways to characterize the
statistical distribution of soak durations were discussed. Regardless of the method of
characterization, it is possible to randomly generate one or more durations for use in a
given test.
Suppose it is decided not to try to describe the database succinctly, but to
retain all 12,073 points. It is possible to generate a random duration with the same
distribution as the original data points by first generating a random integer i with a
uniform distribution between 1 and 12,073. The ith value in the database is the random
duration with the desired distribution.
Suppose, on the other hand, the distribution of durations had been
described by binning the durations in the database. Suppose further that there are k
6-94
-------
bins, and the probability of occurrence of Bin i is p;. To select a bin, first generate a
uniformly distributed random variable x between 0 and 1. Then select Bin 1 if 0 £ x <
pr Select Bin 2 if <; x < p2, etc. It is clear that the probability of selecting Bin i, i= 1
to k, is in accordance with the distribution of the original data points.
The information about the distribution of the durations within bins is not
preserved. It is possible to randomly generate a duration whose distribution is uniform
within the limits of the selected bin, however. This should be adequate, since the bins
are sufficiently narrow for practical purposes. The upper limit of the highest bin must be
assigned somehow; the largest soak duration in the database is one possible choice. The
final randomly selected soak duration, then, has approximately the same statistical
distribution as the original data points.
In any case, assuming an overnight soak is part of each test, it might be
unnecessary to include a second very long soak. Thus, if a soak duration longer than, for
example, 3 hours is generated, this duration can be rejected and another generated. The
combination of an overnight soak, along with one or more soaks of randomly generated
durations restricted to be less than a given value, should be sufficient to represent both
long and short soaks and include a random element.
6.3.2 Signal Processing Analysis
Another advanced analysis approach that could be used effectively to
understand and simulate driving cycles, is the use of signal processing analysis. In the
discussion below, two standard techniques are discussed which could be used jointly to
create a tool to analyze the driving patterns in the database and to produce candidate
driving cycles for consideration for a new Federal Test Procedure.
6-95
-------
6.3.2.1 Conceptual Approach
A signal is any measurable quantity in a system that varies in time. Speech
waveforms, sonar sound pressure levels, seismic waves, and chemical process
concentrations are examples of signals. Because of the long-term interest in these and
other application areas, a comprehensive body of signal processing theory and
application techniques has been developed.
The driving speed of a vehicle as a function of time is a signal. Other
examples of automotive emissions analysis signals include emissions concentrations of
various species, engine RPM, manifold pressure, throttle position, lubricant and coolant
temperatures, and fuel consumption. In this section, only a single channel speed signal
will be considered.
It often helps to use as an analogy a related mature field to suggest useful
techniques. In this section, speech compression for telecommunication will be used as a
model for analyzing vehicle speed, and a neural network-based music synthesis algorithm
will be used as a model for simulating driving patterns. The rest of this section discusses
the issues of vehicle speed representation, model parameter estimation, model validation,
and vehicle speed simulation.
6.3.2.2 Telecommunications Analogy
The direct way to transmit an audio signal over a digital network (such as
the telephone system) is to sample the waveform at a rate based on its bandwidth,
transmit these samples, and pass the samples through a digital-to-analog converter at the
receiver (Figure 6-47A). This was in fact the process used first at the telephone
companies, but they soon discovered that many more lines could be handled with the
same equipment if digital compression and decompression were used (Figure 6-47B).
6-96
-------
a) Telecommunications without data compression.
OS
I
vo
Microphone
Analog-to-
Digital
Conversion
Digital
) '
Compression
i
u
Low Data Rate
r\/\/\y\
Transmit Receive
Digital
Digital-to-
Analog
Conversion
.
Decompression
Speaker
b) Telecommunications with data compression.
Figure 6-47. Telecommunications Signal Transmission
-------
The initial compression schemes considered the audio waveforms as
abstract signals, and generated compact representations based on allowable tolerances on
compression noise levels. This type of compression has been used extensively in digital
telecommunications for several decades.
Because of the rapidly increasing capacity requirements and the new
wireless communications options with limited spectrum space available, improved
compression schemes have been developed. By using a detailed physical model of the
speech generation process (voice box) and the auditory system response (ear), the
intelligent compression schemes represent only the audio content that can be generated
and perceived by humans. By extracting and saving only the essence of the "meaning" of
the signal, the most compact representation of the information is obtained. The total
compression attained is from 700 kb/s (kilobaud/second), the data rate for a high fidelity
audio signal, to about 8 kb/s, the rate for a compressed clearly understandable speech
signal.
The message here for the driving pattern analysis is that the model-based
approach best captures the essence of the information.
For driving, the raw signal is the speed sampled once a second. There is
no insight about the vehicle operation contained in the speed trace itself. But if the
speed trace is transformed to a model-based representation, the vehicle behavior is
expressed in terms of interpreted quantities (such as accelerating from a stop, cruising on
a freeway, slowing for a stop sign, etc.). This is a high-information density form of the
speed trace, just as the model-based audio coding is an information-rich version of the
speech signal. In the next section, the calculation of model-based speed representations
is discussed.
6-98
-------
6.3.2.3 Vehicle Speed Representation
Figure 6-48 shows the detail from a segment of sampled speed. The x's on
the graph denote the sample values. This is the lowest level of representation of the
vehicle speed, and the "meaning" of the vehicle behavior is not explicit. Parameters that
define the sampling process are the sampling rate (fast enough to capture rapid speed
changes), amplitude resolution (to represent small speed fluctuations), and dynamic
range (to cover the full range of possible speeds).
The next level is to represent the speed signal as a sequence of "states".
Each state, or driving segment, is described by the type of state and several numerical
parameters.
EXAMPLES OF DRIVING STATES
State
Parameter
Engine off
Idle
Acceleration
Cruise
Deceleration
Duration
Duration
Initial speed, final speed, duration, sigma
Initial speed, final speed, duration, sigma
Initial speed, final speed, duration, sigma
The parameter "sigma" is a measure of the variation of the acceleration
during the segment. Routines could be easily written to transform any driving cycle to a
sequence of such states. The number of states generated would vary, depending on the
fidelity requested. Higher fidelity would require more states.
Representing the speed as a sequence of abstract states has several
advantages for driving cycle analysis and simulation. The information is compressed,
especially when the fidelity requirements are not too stringent. Events related to
6-99
-------
Figure 6-48. Example Driving Segment
6-100
-------
emissions, such as engine startup, engine off time, and rapid acceleration, are
represented explicitly.
The third level of representation is based on a physical model of driving,
including such factors as vehicle types and responses, driver behavior types, road
conditions (speed limits, stop signs, hills, etc.), and traffic levels. Just as raw speech
signals are transformed into phonemes, sentences, and meaning with automatic speech
recognition algorithms, the raw speed signals can be transformed into driving segments
and trips by automatic driving recognition algorithms.
The advantage of using a representation based on a deep knowledge of
vehicle and driving characteristics is that the information density is especially high and in
a form useful for analysis, insight, and simulation. Predicting the effect of changes in
vehicle design or traffic laws on the driving pattern is straightforward and accurate. The
disadvantage, of course, is the amount of effort required to develop and validate the
model-based representation.
6.3.2.4 Neural Network Simulations
Simulation methods, based on either driving theory or measurements, can
be used to generate example speed signal traces similar to real-world driving. If the
process is well understood, the simulation can be based on a theoretical model. When a
large amount of data is available, as is the case with driving, the simulation algorithm
can be trained to behave like the example data.
There are several possible reasons for building a simulation of driving.
The simulation system will be a source of realistic signals for research and testing. By
varying model parameters, the effect of controllable changes to the driving characteristics
(such as different traffic laws, new vehicle responses, etc.) can be predicted. And
perhaps the most important benefit will be the increased understanding of the driving
6-101
-------
process resulting from the development of the simulation and the validation of the
results.
Recently developed neural network techniques have been successfully
applied to solve complex pattern recognition and control problems. Based loosely on
nervous system function, neural networks are most suitable for complex problems for
which a large amount of data is available. Generating vehicle speed traces characteristic
of a particular type of driver and driving conditions is this class of problem. Neural
networks have been trained to solve a similar problem, learning a musical grammar from
examples and composing music in the style of the examples. Two neural network
approaches to this music problem are described briefly below, along with parallels to the
driving modelling problem.
The first paper (20) describes a neural network program that can learn
melodic passages and counterpoint style from examples and then generate new passages
in the same style. The author trained the network with the three-part Inventions of J. S.
Bach and generated music that was "smooth, continuous, and pleasant."
Kohonen listed the main problems in computer music, which all have their
analogies in driving models. First, it is difficult to automatically find the start and stop
of possible themes. In driving, it is difficult to unambiguously define the start and end of
driving segments, such as trips, acceleration sections, cruising, etc. Second, it is difficult
to maintain a certain "style" over a passage and switch to a new one at the transition to
the next passage. This might correspond to the transition from one style of driving
versus another ("easy Sunday drive vs. reach the restaurant before closing time" or
"freeway" vs. "residential"). Third, special rules are required in the music generator to
prevent bad melodic turns, but even with these rules the resulting music still isn't always
beautiful. The corollary for driving is that special rules will be needed to prevent
unphysical speed traces (e.g., negative speeds, or accelerations beyond the vehicle's
limits), but some unrealistic sections will continue to appear periodically.
6-102
-------
The heart of Kohonen's approach is to address these difficulties with a
"dynamically expanding context". The number of past notes used to determine the next
note varies with the location in the piece, and this optimum length of context is
determined from examples separately for each note sequence. The grammar is stored as
a set of matrix of coefficients that define a set of production rules. The knowledge of
the neural network, a form of associative memory, is stored in the weights. The network
is trained by applying musical sequences to the inputs (in this case Bach's Inventions),
and adjusting the weights to learn the musical patterns.
New, partly random, music sequences are generated by using these stored
production rules to compute the notes. Often several different production rules apply in
a given situation, and a particular one can be selected randomly to introduce variety in
the melody. Musically knowledgeable listeners perceive the resulting music as
mechanical but in the style of Bach.
If a similar technique is applied to driving patterns, the variable length of
context supports learning about different types of trips with different time scales:
backing up in the driveway, traversing city blocks, and extended segments of freeway
driving. Musical pieces are described as sequences of notes, of lengths of several
hundred to several thousand notes. For training, it is advantageous to transform the
speed signal samples to "feature" sequences (such as acceleration, deceleration, cruise,
etc.), as discussed above, to reduce the input vector size.
Another neural network for learning musical patterns was trained from the
music of George Frederick Handel (21). This author used the ART 3 architecture to
design a neural network capable of learning the varying-scale embedded temporal
patterns of chords that characterize harmonic syntax. This paper illustrates how carefully
designed neural network algorithms, together with large amounts of training data, can
recognize and learn the finer detail structure of the input sequence. Since vehicle speeds
6-103
-------
also have correlation patterns of varying length, a similar technique could be used to
improve the accuracy of driving simulations.
6.3.2.5 A Zeroth Order Signal Processing/Neural Network Driving Pattern Model
Vehicle driving behavior can be visualized in different ways. The tradi-
tional way of viewing it is as a series of trips during which the driver accelerates the
vehicle to a desired speed and then decelerates when he gets to his destination. The
driver accelerates his vehicle away from stop signs and stop lights. The basic concept of
this traditional model is that the natural state of the vehicle is at rest with the engine off.
Any model that is used must explain why the vehicle is not at rest with the engine off.
To do this, reasons must be categorized, explained, or modeled for all accelerations and
speeds other than zero.
Any conceptual models can be used to understand and analyze driving
pattern behavior, but some of the models may have advantages over others in terms of
the simplicity of the mathematics. One that has been attractive to the authors in this
study has been based on the concept that the natural or base state of the vehicle is that
it is running at a constant and high speed. The rationale for this base state is that the
vehicle was manufactured to be driven and that drivers want to go as fast as possible to
reach their destination as soon as possible, subject to the driving conditions. The only
reason that the vehicle would not be driven at the constant high speed is that the vehicle
encounters an "obstacle" for which the driver judiciously causes the vehicle to slow down.
After the obstacle has been passed, the speed of the vehicle again increases to its natural
high constant speed.
One of the attractive features of this obstacle model is that there are only
a few types of obstacles which are needed to explain a decrease in the speed of the
vehicle. There are basically five types of obstacles that need to be considered:
6-104
-------
• Natural fluctuations in speed;
• Objects to be avoided;
• Curves and turns in the road;
• Traffic controls; and
• The destination.
Even though the driver would like to maintain his vehicle at a constant
high speed, because of natural fluctuations in the vehicle power plant operation and
small grade changes in the road, there will always be small speed changes. While this
type of operation can be maintained on straight, level roads, usually other traffic is in the
vicinity and the driver must avoid this other traffic. This causes his vehicle to need to
slow down. For safety, the driver must slow down for objects, such as wet pavement or
an animal running in front of his vehicle.
In this study, the authors instrumented their own vehicles and kept a
written log of where and when the vehicle was driven for each trip. Examination of the
speed profile, in comparison with the written log, showed that the speed of the vehicle
decreased when the vehicle went around curves in the road or made turns at inter-
sections, even though no traffic controls forced the driver to slow the vehicle down.
If traffic controls are present, these have an obvious effect on the driving
pattern of the vehicle. Speed limit signs limit the upper speed of the vehicle to some
value. Stop signs cause the driver to slow the vehicle down; in most cases, it is not a
stop, but a rolling stop. Stop lights cause vehicles to stop when they are red and usually
to slow down when they are green. Finally, reaching the destination, where the vehicle is
shut down, is the ultimate obstacle in this model.
6-105
-------
It could be that these five categories of obstacles do not describe all events
in driving patterns; however, they probably describe a large percentage of them.
Another advantage of this obstacle model is that each of the obstacles that cause a
reduction in the speed of the vehicle may do so in a characteristic way. Thus,
examination of the speed data in the driving pattern database, using appropriate pattern
recognition software, may be able to identify, for a large fraction of the obstacles, the
nature of the obstacle that was encountered by the driver. A cluster analysis of the
obstacle characteristics could be performed.
The speed vs. time profiles for each of the encounters with any obstacle
may have similarities that can be described by a relationship between speed and time.
Different obstacles would be described by different parameters in the relationship. The
result would be that an obstacle model could describe all changes in vehicle speed with
one or a few relationships. It would be the parameters that change from obstacle to
obstacle.
One such obstacle model that could be used to describe the encounters of
a driving vehicle is analogous to a pinball machine. This special pinball machine has
rubber bumpers spread throughout its frictionless playing field. The steel ball represents
the vehicle. At the beginning of the trip, the ball is shot onto the playing field. It travels
for a distance at a high constant speed, until it encounters a rubber bumper. If it hits
the rubber bumper head-on, the speed of the steel ball slows rapidly to 0 mph and then
accelerates again as it bounces off the rubber bumper. If the ball hits a rubber bumper
in a grazing fashion, it does not stop but slows down to a degree, depending on the angle
of incidence with the bumper. If the playing field is filled with many rubber bumpers
with different softnesses of rubber and with different populations, the speed of the steel
ball would produce a very complex-looking speed vs. time profile as it moved through
the playing field. The softness of the rubber in the pinball bumpers is analogous to the
power/weight ratio for vehicles for accelerations and braking efficiency for decelerations.
6-106
-------
In a pinball machine, an understanding of the motion of the ball depends
on an understanding of the population distributions of rubber bumpers, their softnesses,
and locations. These parameters can be determined on a pinball machine by taking
measurements of the speed of pinballs during trips on the playing field. In an analogous
manner, the characteristics of obstacles that vehicles encounter on the road that cause
them to slow down and speed up can be determined by examining the speed profiles of
many vehicles on the road. These measurements have been done in this study. By
analyzing these measurements in an intelligent fashion, it should be possible to describe
the obstacles in enough detail that simulated driving patterns can be created which have
all of the characteristics of the driving patterns measured in this driving pattern study.
Developing such a model may or may not be a simple task; however,
through the use of a combination of signal processing analysis and artificial neural
networks, it seems that it is possible.
A zeroth order model based on a body (the vehicle) striking a rubber wall
(the obstacle) at an angle is described below. The interaction of the body with the
rubber wall is modeled using Hooke's Law for the normal component to the wall and the
minimum speed attained at the collision to be equal to the speed of the body parallel to
the rubber wall. This model produces sines and cosines for the speed profiles for a
collision perpendicular to the rubber wall and a more complex equation for speeds
striking the wall at an angle. To demonstrate how the signal processing analysis and
neural network could be used to learn about driving pattern behavior and then to
simulate driving patterns, an example driving segment is presented in Figure 6-49.
In the example driving segment (Figure 6-49), the dashed line is the
measured speed and the solid line is the speed predicted by a 17 segment obstacle model
fit. The segments are defined by 10 obstacles, which were encountered at the times
shown by the arrows. Each segment is defined by three parameters: initial speed, final
speed, and duration (Table 6-7). The fourth parameter in the table (k), is computed
6-107
-------
80 100
Seconds
Figure 6-49. Obstacle Model for Example Driving Segment
6-108
-------
Table 6-7
Obstacle Description Parameters
Speed
Segment
Speed Fitting Parameters
Calculated
Collision
Softness
k
Beginning Speed
(m/s)
Ending Speed
(m/s)
Segment Duration
(s)
1
0.6
8.1
10
0.15
2
8.1
7.9
5
0.31
3
8.0
8.3
9
0.17
4
7.9
5.7
4
0.39
5
6.7
10.9
5
0.31
6
10.6
1.6
9
0.17
7
2.4
15.9
13
0.12
8
16.0
17.4
7
0.22
9
17.3
16.0
9
0.39
10
16.2
16.7
4
0.39
11
16.6
15.7
6
0.26
12
15.9
17.3
12
0.13
13
17.3
16.9
6
0.26
14
17.0
19.2
9
0.17
15
19.2
17.9
9
0.17
16
18.0
18.9
11
0.14
17
18.6
0.0
18
0.08
6-109
-------
from the other three parameters and is analogous to the spring constant of harmonic
motion.
Once the physical modeling parameters are determined for each segment,
the parameters for all obstacles in the database are used to train a neural network. This
second step in the model has not been done for this example, but conceptually it
completes the development of the driving pattern model.
The physical modeling parameters form a condensed representation of the
driving information. In the example, the original 147 time samples are represented by 17
segments defined by three parameters each, for a data compression ratio of about 3 to 1.
The physical model can be iteratively refined, using large measured driving data sets for
testing and validation, to improve the both data compression and model accuracy. The
value is not principally in the data compression, but because the parameters capture the
essence of the driving information in a compact form. The amount of compression is a
measure of the accuracy and generality of the model. Physical models are useful for
driving analysis tasks, such as determining the differences between different types of
drivers or different vehicle power-weight ratios. The model representation is also useful
for comparing driving histories, as would be required, for example, for validating
simulated driving cycles.
Physical driving models are especially useful for generating simulated
driving cycles. The simulator merely needs to select driving parameters according to the
measured statistics of the type of driving desired, and the physical model will accurately
fill in the fine detail of the speed vs. time traces. The simulator algorithms could be
based on any trainable method, such as neural networks, fuzzy logic, rule-based expert
systems, or traditional statistics.
6-110
-------
Although the simple obstacle model represents the example data section
fairly well, some straightforward model enhancements could improve the model
performance significantly.
First, the obstacle model uses the same functional form for acceleration
and deceleration segments. But we know that different physical systems (braking and
engine power) are involved, and the speed responses would be better represented by a
model which incorporates the characteristics of these different systems. The model
should also require continuity of the speed and acceleration at the junctions between
segments. The obstacle model in the example allows discontinuity of acceleration at
these points.
Next, some form of multi-scale analysis should be supported by the model.
The user should be able to specify the accuracy required, and the model would adjust
the tightness of fit and the number of parameters required accordingly. The few
parameters required for a "loose" fit describe the general shape of the speed curve, and
the additional parameters added for tighter tolerance requirements describe the detail of
speed fluctuations.
By choosing the right mathematical forms for the model, the parameter
estimation process can be fully automated. This allows the use of the massive amounts
of collected data to test and refine both the functional form and the parameter statistics
of the physical models.
6.4 Options for Formulating Driving Cycles
The ultimate goal of this driving pattern study is to evaluate the current
certification test procedure, the FTP. These data will be used to determine whether the
FTP is adequate, needs to be updated, or needs to be replaced. The current test
6-111
-------
procedure specifies the speed of the vehicle versus time, and we anticipate that any new
procedure that may replace it would use a similar approach.
6.4.1 Two Possible Bases for a Certification Test Procedure
A new certification test procedure can be developed using one of two
bases. The cycles can be representative of today's:
• Driving patterns; or
• Emissions patterns.
Each basis has advantages and disadvantages, as discussed below.
A driving cycle based purely on driving patterns would be representative of
the driving pattern of the vehicle population. Emission measurements taken while
vehicles are driven in traffic indicates that such a cycle would spend most time testing
driving features that produce low emissions and a small amount of time testing high
emission driving features. During certain episodes, for example, during cold start
operation or during heavy acceleration, HC and CO emissions are high. However, these
high emission events occur during a small fraction of the total driving time. This means
that a driving cycle representative of vehicle population driving patterns will be
inefficient at measuring the performance of the vehicle emission control systems during
high emissions episodes, although the cycle is accurately measuring the emissions
performance of the vehicle in a cycle representative of real world driving.
Alternatively, a certification test procedure based on the emissions pattern
of the vehicle population will reflect the range of driving in the real world but will be
weighted by the response of the vehicle emissions to those driving features. Because of
this, the emissions-based driving cycle will contain more high emission driving features
6-112
-------
and fewer low emission driving features. Thus, the procedure would be efficient at
measuring the high emissions behavior of vehicles.
An example will help clarify the important distinction between a driving
pattern-based certification test procedure and an emissions pattern-based certification
test procedure. Suppose that analysis of the driving pattern data and a separate study of
the relationship between emissions and driving indicates that the emissions rate of
hydrocarbon can be explained by just two kinds of driving modes: Type A and Type B.
Suppose that it is found that vehicles spend 95% of their driving time in Type A driving
but generate only 20% of the total fleet hydrocarbon emissions in this mode. Suppose
that vehicles spend only 5% of their driving time in Type B driving but generate 80% of
the total fleet hydrocarbons in this mode.
The driving based certification test procedure based on this hypothetical
information would be a cycle with 95% of the time in the A mode and 5% of the time in
the B mode. The emissions-based certification test procedure would be a cycle with 20%
of the time in the A mode and 80% of the time in the B mode. This second procedure
emphasizes the time spent on high emissions features.
A driving cycle based purely on driving patterns will need to be updated
only when the driving patterns of the vehicle population change. Changes may occur
relatively slowly over perhaps twenty years. A driving pattern-based test procedure will
also continue to representatively measure the emissions of vehicles although the vehicle
emission control technology and fuel types may change dramatically in the next several
years.
A driving cycle based on emissions patterns depends on the emissions
behavior of current vehicles, which in turn depends on the type of fuel and emission
control technology used when the current vehicles were being certified. Thus, for
example, a new driving procedure developed in the next few years would be based on the
6-113
-------
emission characteristics of 1992 vehicles. These vehicles are, of course, creatures of the
current Federal Test Procedure, since they have been designed to pass that procedure.
As vehicle manufacturers continue to develop more advanced control technologies and
as new types of fuels become available in the marketplace, a driving cycle based on
emissions behavior of 1992 vehicles may become a poor measure of the emissions
behavior of the latest vehicles. Consequently, the new driving cycle may become
obsolete in a short time. A continually revised certification test procedure that responds
to changes in vehicle technology and fuels, however, will mean that the rules for
manufacturing a new vehicle will be continually changing. Manufacturers will find such
procedural changes difficult to quickly respond to, given the long lead times needed to
integrate a new control system with a future model year vehicle. Also, EPA would need
to update the certification test procedure regularly. In addition, there is no guarantee
that this circular process would converge toward lower vehicle emissions and a simpler
certification test procedure.
Another disadvantage of the emission-based test procedure is that the
desired weighting of the high and low emitting driving features may not be the same for
HC, CO, and NOx emissions. Road emissions data indicates that HC and CO emissions
increase greatly during cold start operations and heavy accelerations, but the response of
NOx emissions to driving may be different.
The choice of whether to use a driving pattern-based or an emissions
pattern-based test procedure is difficult because which will be more effective depends to
a large extent on the speed and degree of changes that may occur in emission control
systems, fuel composition, and emissions of concern. If changes are slow and small, and
if the new procedure can be implemented rapidly, the emissions-based procedure will be
more effective for several years. If changes happen quickly and are large, then the
emissions-based procedure will remain effective for a shorter time.
6-114
-------
6.4.2
Format Options for a New Test Procedure
The final development of the new certification test procedure or updating
of the current FTP will depend on the basis of the new procedure and the format that
the new procedure might have. As discussed in Section 6.4.1, the certification test
procedure may focus on driving patterns or emission patterns. The test procedure could
be developed in a straightforward manner as a direct consequence of the advanced
statistical analysis techniques or the signal processing data analysis techniques. In the
discussion below, options for the revised test procedure are discussed.
Several different certification test procedure formats can be envisioned for
an updated or new procedure. These might include procedures based either on driving
patterns or emission patterns. The formats include:
• Current FTP;
• Revised FTP (Bag 4);
• New deterministic cycle;
• Multiple deterministic cycles; and
• Multiple stochastic cycles.
After the driving pattern data are fully analyzed, it may be decided that the
current FTP is a good representation of today's driving and therefore does not need to
be revised. However, the current FTP is believed not to cover the full range of
accelerations present in the vehicle population. Since in today's technology vehicles,
heavy accelerations are known to produce emissions, the FTP may be underestimating
the emission potential of today's new vehicles. One approach to correcting this situation
is to add a Bag 4 to the end of the current FTP. This fourth bag would contain the
higher accelerations and other emission features that need to be added to the FTP to
6-115
-------
make it more representative of actual emission patterns. The driving features that would
be included in Bag 4 cannot include all of the driving pattern features that would make
the FTP representative of driving patterns in the population. If this is done, most of the
driving features in Bag 4 could not contribute significantly to emissions produced by the
FTP cycle. Since the features that should be included in Bag 4 are based on emissions,
it is imperative that the relationship between emissions and driving be known. This
would have to be determined in a separate study or be based on existing information.
If the signal processing approach is used to analyze the data, and if the
driving pattern descriptors are evaluated for their contributions to emissions, the current
FTP can be evaluated to determine the types of driving features that need to be added
to make it more representative of emissions patterns of the vehicle population.
Another approach is to replace the current FTP with a new deterministic
cycle (that is, a cast-in-iron cycle). A most likely deterministic cycle could be produced
using a simulation based on the advanced analysis of the data. If this new cycle is based
on emission patterns, the most likely cycle will emphasize the high emission driving
features of the vehicle population. However, if this cycle is based on driving patterns,
the most likely cycle will not include very much high emission feature content, since the
test procedure would be representative of population driving.
An alternative approach is to use multiple deterministic cycles. Such a
battery of cycles could be used to test the vehicles near the extremes of driving behavior
measured in the driving pattern study.
Rather than deterministic cycles, which are fixed, multiple stochastic
(randomly generated) cycles could be used to certify new vehicles. Such cycles could be
generated using a simulation based on the results of the advanced statistical or the signal
processing analysis of the driving pattern data.
6-116
-------
Randomly generated cycles would have a variety of driving features; the
severity of each feature would be randomly selected from the range of severities found in
the driving pattern study. For a new vehicle to pass certification, it would have to pass a
certain fraction of a certain number of randomly generated driving cycles. A computer
program that generates the cycles would be made available to anyone for testing
purposes. The fraction of stochastic cycles which must be passed and what determines a
pass would have to be decided using statistical and probabilistic techniques.
The advantage of using the stochastic approach to develop a driving
pattern based test procedure is that the approach will work regardless of the emission
control technology, the fuels being used by the vehicles, or the emissions of concern. In
addition, the stochastic approach would be effective because the procedure would
encourage manufacturers to develop emission control systems that could pass most
randomly generated cycles. EPA's testing of vehicles to be certified might be time-
consuming, but the development of an emission control system would not require the
testing of numerous stochastic cycles. In the development of a vehicle emission control
system, the manufacturer might choose to test the vehicle only for those features where
emissions are high. Such a certification test procedure would be good as long as the
driving pattern data used to develop the driving pattern descriptors and their distribution
of values are good.
6-117
-------
7.0 REFERENCES
1. DeFries, T.H. and R.F. Klausmeier, "Evaluation of Driving Pattern
Measurement Techniques," Radian Corporation, DCN 91-298-018-01-02,
18 July 1991.
2. "Federal Test Procedures, Part 86 of Title 40." Code of Federal
Regulations (40 CFR 86).
3. DeFries, T.H. and S. Kishan, Driving Modes Quality Assurance Project
Plan, Revision 1. Radian Corporation, DCN 92-254-036-90-05, 31 January
1992.
4. DeFries, T.H. and G.F. Baker. Instrumented Vehicle Pilot Study of In-Use
Driving Behavior in Spokane. Washington. Radian Corporation, DCN 92-
245-130-02, 24 January 1992.
5. Horowitz, A.D. "Automobile Usage: A Factbook on Trips and Weekly
Travel." GMR-5351, General Motors Research Laboratories, Warren,
Michigan, 2 April 1986.
6. Maurin, M. and J. Crauser. "The Kinematic Sequences, an 'Atomistic'
Approach to Automobile Travel and the Effects of Travel," Recherche
Transports Security. English Issue No. 5, Date unknown.
7. Smith, M. Development of Representative Driving Patterns at Various
Average Route Speeds. Scott Research Laboratories, Inc., Report No.
EPA-450/3-76-023, February 1974.
8. "Should Europe Adopt a Modified FTP?" Automotive Engineering. V 86.
Warrendale, Pennsylvania, 6 June 1978. pp. 50-53.
9. Milkins, E. and H. Watson, "Comparison of Urban Driving Patterns." In:
Proceedings of the Motor Vehicle Technology: Progress and Harmony.
Second International Pacific Conference on Automotive Engineering.
Society of Automotive Engineers (Australasia), Tokyo, Japan, November
1983. pp. 735-745.
10. Watson, H.C. "Influence of Vehicle Driving Patterns on Localized Urban
Emissions Sources." In: Proceedings of the Automobile Engineering
Meeting. Society of Automotive Engineers, Detroit, Michigan, May 1973.
7-1
-------
11.
Harkins, J., M. Smith, and M.J. Manos. Vehicle Operations Survey
Volume I. Scott Research Laboratories, Inc., SRL-2922-13-1271; CRC-
APRAC-CAPE-10-68-9, San Bernardino, California, 17 December 1971.
12. Johnson, T.M., D.L. Formenti, R.F. Gray, and W.C. Peterson.
"Measurement of Motor Vehicle Operation Pertinent to Fuel Economy."
In: Proceedings of the Automotive Engineering Congress and Exposition.
Society of Automotive Engineers, Detroit, Michigan, February 1975.
13. Stivender, D.L. and Y.G. Kim. "Dynamic Computer Techniques for
Vehicle Emission Development." In: Proceedings of the Automotive
Engineering Congress. Society of Automotive Engineers, Detroit,
Michigan, January 1972.
14. Deshpande, G.K. "Development of Driving Schedules for Advanced
Vehicle Assessment." In: Proceedings of the International Congress and
Exposition. Society of Automotive Engineers, Detroit, Michigan, February
1984.
15. Bullock, K.J. "Driving Cycles." In: Proceedings of the Second Conference
on Traffic. Energy, and Emissions. Society of Automotive Engineers
(Australasia), Melbourne, Australia, May 1982.
16. Andre, M. "In Actual Use Car Testing: 70,000 Kilometers and 10,000
Trips by 55 French Cars under Real Conditions." In: Proceedings of the
International Congress and Exposition. Society of Automotive Engineers,
Detroit, Michigan, February 1991.
17 Williamson, H., T.H. DeFries, R.F. Klausmeier, J. Lange, and L. Collins.
Preliminary Analysis of Data Collected in EPA's Operational
rharacteristics Study. Radian Corporation, DCN 92-298-018-01-05,
2 October 1992.
18. Hartigan, J.A. Clustering Algorithms. Wiley-Interscience, 1975.
19. Gumbel, E.J. Statistics of Extremes. Columbia University Press, 1958.
20. Kohonen, T. A Self-Learning Musical Grammar, or "Associative Memory
of the Second Kind." In: Proceedings of International Joint Conference on
Neural Networks. Washington D.C., June 1989, pages 1-1 to 1-5.
21. Gjerdingen, R.O. "Learning Syntactically Significant Temporal Patterns of
Chords: A Masking Field Embedded in an ART 3 Architecture." Neural
Networks. Vol 5, pp 551-564, 1992.
7-2
-------
APPENDIX A
SAMPLE QUALITY CHECKING PLOTS
-------
Figure A-l
19MAR92:04:53 21MAR92:12:26 23MAR92:20:00 26MAR92:03:33 28MAR92:11:06 30MAR92:18:40 02APR92:02:13
Date
fu631495.3rd
-------
Figure A-2
DATE=21MAR92
10:26:40 10:28:20 10:30:00 10:31:40 10:33:20 10:35:00 10:36:40 10:38:20
Time
fu631495.3rd
-------
Figure A-3
DATE=31MAR92
Time
fu631495.3rd
-------
Figure A-4
19MAR92:04:53 21MAR92:12:26 23MAR92:20:00 26MAR92:03:33 28MAR92:11:06 30MAR92:18:40 02APR92:02:13
Date
fu631495.3rd
-------
Figure A-5
10000 20000 30000 40000
Cumulative Engine-On Time (s)
fu631495.3rd
-------
40 H
30
Figure A-6
(/)
"O
-------
Figure A-7
40-
10000 20000 30000 40000
Cumulative Engine—On Time (s)
fu631495.3rd
-------
Figure A-8
4000-
3000-
Q_
cr
v 2000
c
c
Ld
1000
0-
'[ I I I I I I I I I | I f I I I 1 I I I | I > I I I > I I I | I I I I I I l 1—I p
0 10000 20000 30000 40000
Cumulative Engine —On Time (s)
fu631495.3rd
-------
Figure A-9
100-1
90
O
CL 80
-------
Figure A-10
(J)
> 3
\
£ 2
c
o
• «
o
1
0
_
-------
Figure A-ll
3000-1
2000-
CO
'cr>
c
LlI -2000-
¦3000 —)—|—,—,—,—,—,—,—,—j—,—,—,—,—,—,—,—,—,—(—,—,—,—,—,—,—,—,—1—(—i—, , .,—,—,—(—1—,—r
0 10000 20000 30000 40000
Cumulative Engine-On Time (s)
fu631495.3rd
-------
50
40
30
20
10
0
10
20
30
40
50
60
Figure A-12
10000 20000 30000 40000
Cumulative Engine —On Time (s)
fu631495.3rd
-------
Figure A-13
i i i i |—i i i i i—i i i—i—| i i i—i—i—i—i—i—i—p
20 30 40
Speed (m/s)
fu631495.3rd
-------
6
5-
if)
C/)
v
3
E
2-
c
o
o
-------
Figure A-15
4000 -j
3000
Q_
cr
O 2000
c
"cn
c
1000
0
100
Manifold Abs. Pressure (kPa)
fu631495.3rd
-------
0.013-j
0.012
0.011
£ 0.010
Cl
*- 0.009
\ 0.008
0.007
^ 0.006
CL
0.005 -
"S 0 004 ¦
a>
CL 0.003
CO
0.002
0.001
0.000
1
0
Figure A-16
I I I I I I I I
20
Speed (m/s)
30
-r—1—r
40
fu631495.3rd
-------
APPENDIX B
DATA ANALYSIS PLOTS AND TABLES
-------
Table B-l
DBS
SITE
SAMPLE
VIN
MAKE
1
Balt-Ross
B002
1FTCR10A1MUC48371
ford
2
Bait-Ross
B005
3B3XA4633MT034156
Dodge
3
Bait-Ross
BOOS
1G1LT64W8KY218187
Chevrolet
4
Bait-Ross
B010
7V1AX885XJ1793960
vol vo
5
Balt-Ross
B011
1B3XG24K5KG180041
Dodge
6
Bait-Ross
B016
1G1FP21S3KL104669
chevrolet
7
Bait-Ross
B017
YS3AL75L8M7006092
saab
8
Bait-Ross
B019
1G1FP21S1JL193186
chevrolet
9
Bait-Ross
B037
2P4FH413XJR662190
pi ymouth
10
Bait-Ross
B039
1P3BK.4602KC469869
Plymouth
11
Bait-Ross
B042
1G6AD6983D9266654
cadi 11ac
12
Bait-Ross
B044
1G2HZ54C2KW215880
Portiac
13
Bait-Ross
B050
1GCEG25HXC7125742
Chevrolet
14
Salt-Ross
8052
1HGCB7661MA049561
handa
15
Bait-Ross
8053
1G3NT14DXKM255041
01 dsmobi 1 e
IB
Bait-Ross
B055
JM1GD2224L1810193
Mazda
17
Bait-Ross
B058
1G2AB2705E7332521
pontiac
18
Bait-Ross
B059
1G1JF11W2K7180967
Chevrolet
19
Bait-Ross
B062
1FAPP14J2MW2G9094
Ford
20
Bait-Ross
B066
1YVGD31B5M513870B
MAZDA
21
Bait-Ross
B072
1FTCR10A10UK81004
ford
22
Bait-Ross
B074
1MEPM6042LH679380
Mercury
23
Bait-Ross
B081
1G1JC1113KJ171405
Chevrolet
24
Bait-Ross
B084
1G2AB2707 E7250564
PONTIAC
25
Bait-Ross
B086
1FAPP36X9KK20166B
ford
26
Bait-Ross
B088
1GCDC14ZBKZ2U565
chevrolet
27
Bait-Ross
B098
1GTBS14R2GS533452
gmc
28
Bait-Ross
B099
1G3HY14C3KW3H671
01 dsmobi 1 e
29
Bait-Ross
B105
JT4RN50R7G0173173
toyota
30
Bait-Ross
B113
1N4GB22 53KC735159
Ni ssan
31
Bait-Ross
B119
1G1JF14T4H7133492
Chevrolet
32
Bait-Ross
B123
TE725009B80
toyota
33
Bait-Ross
B124
1B3BP48D5KN589752
dodge
34
Bait-Ross
B130
1FABP44E2KF117098
ford
35
Bait-Ross
B139
1G4AL1937FG463396
buick
36
Bait-Ross
B140
1MEPM36X2KK620959
Mercury
37
Bait-Ross
8143
JT2RA44C6B00O1018
toyota
38
Bait-Ross
B144
JE3CU36X6KU035123
Eagle
39
Bait-Ross
B150
1B3YA44K6JG448546
dodge
40
Bait-Ross
B153
1GNCT18Z3K0152538
Chevrolet
18:13 Friday, August 7, 1992
MODEL
YEAR
TPL
TPH
TR
ranger
1991
A
Spirit
1991
30
1913
A
Corsica
1989
0
100
A
240dl
1987
A
Daytona
1989
512
2075
A
camaro
1989
A
900
1991
M
camaro
1988
A
voyager
1988
A
reliantk
1989
A
sedandeville
1983
A
Bonnevi11e
1989
0
100
A
c20van
1982
A
accord
1991
A
Calais
1989
0
100
A
626
1990
250
2173
M
2000
1984
A
cAvelierz24
1989
A
Escort
1991
M
mx-6
1991
M
ranger
1989
M
Cougar
1990
426
2316
A
cavalier
1989
A
SUNBIRD
1984
A
tempo
1989
A
si lverado
1989
A
sl5
1986
A
88
1989
0
100
A
pickup
1987
M
Sentra
1989
665
4298
M
Cavalier
1991
0
100
A
corollawagon
1980
M
shadow
1989
A
mustanglx
1989
A
century
1985
A
Topaz
1989
364
2153
A
eelica
1981
M
Surnnit
1989
262
2512
A
daytona
1988
A
blazer
1989
A
-------
Table B-l (Continued) 18:13 Friday, August 7, 1992 2
DBS
SITE
SAMPLE
VIN
MAKE
MODEL
YEAR
TPL
TPH
TF
41
Balt-Ross
B156
1GCFC24H1KE115676
Chevrolet
silverado2500
1989
A
42
Balt-Ross
B163
JHMCA5386JC052224
honda
accordlxi
1988
M
43
Balt-Ross
B164
1B3BD31D6FG241191
CHRYSLER
ARIESK
1984
A
44
Balt-Ross
B165
2B4GK55R7MR201092
Dodge
Caravan
1991
413
1955
A
45
Balt-Ross
B167
1YVGD22B0M5125967
Mazda
626
1991
246
2159
A
46
Balt-Ross
B169
1G4XB69R0FW454358
bui ck
skylark
1985
A
47
Balt-Ross
B178
JH4DA3458KS013350
accura
integra
1989
A
48
Balt-Ross
B184
1G4AH51R1HT438622
bui ck
century
1987
A
49
Balt-Ross
B190
JT2EL43B2M0027569
toyota
tercel
1991
A
50
Balt-Ross
B191
1GCDC14H3ME100424
Chevrolet
1500
1991
A
51
Balt-Ross
B193
1GCDC14Z6KE135013
Chevrolet
scotsdalel500
1989
A
52
Balt-Ross
B195
JHMBA7431GC066566
honda
accord
1986
A
53
Balt-Ross
B206
2MEBM75F1KX665883
MERCURY
GRANDMARQUIS
1989
A
54
Balt-Ross
B208
JM1FC3312G0121078
MAZDA
RX-7
1986
M
55
Balt-Ross
B210
JF1AC42B2KC210379
subaru
dl
1989
A
56
Balt-Ross
B233
JN1GB21SXKU537381
NISSAN
SENTRA
1989
A
57
Balt-Ross
B236
1W19J9B567918
CHEVROLET
MALIBU
1979
A
58
Balt-Ross
B243
YV1AX8852H1723847
VOLVO
240DL
1987
M
59
Balt-Ross
B247
2B4FK5131JR676116
DODGE
CARAVAN
1988
A
60
Balt-Ross
B255
1C3BC59KXHF330954
CHRYSLER
T0WNANDC0UNTRY
1988
A
61
Balt-Exet
B268
1G1AW35K5CR225618
Chevrolet
Maii bu_Classi c
1982
A
62
Balt-Exet
B273
1FTCR11T1JUE26260
ford
Ranger
1988
A
63
Balt-Exet
B275
JT2AL31G7E0225492
TOYOTA
TERCEL
1984
M
64
Balt-Exet
B286
CRN1498264964
CHEVROLET
LUV
1979
M
65
Balt-Exet
B287
1237J9B412061
CHEVROLET
M0NTECARL0
1979
A
66
Balt-Exet
B297
1MEBP9237GH654018
MERCURY
COUGAR
1986
A
67
Balt-Exet
B302
1G1LT54T2LY160319
Chevrolet
Corsica
1990
100
A
68
Balt-Exet
B314
1G3CX69B9G4332506
0LDSM0BILE
98
1988
A
69
Balt-Exet
B315
1FTEX15H2GKA94158
FORD
F150
1986
A
70
Balt-Exet
B317
8G87H169118
FORD
THUNDERBIRD
1978
A
71
Balt-Exet
B319
1FABP259XHW11G408
FORD
ESCORT
1987
A
72
Balt-Exet
B329
1G1GZ11G2HP118178
CHEVROLET
M0NTECARL0
1987
A
73
Balt-Exet
B337
JT2MX62EX80014430
toyota
cressi da
1981
A
74
Balt-Exet
B338
1G6A0478XC9150406
CADILLAC
C0UPDEVILLE
1982
A
75
Balt-Exet
B344
1Y1SK5167LZ072182
GEO
PRIZM
1990
A
76
Balt-Exet
B351
1G4AJ47A8EH477206
buick
rega
1984
A
77
Balt-Exet
B354
1FAPP36X0KK176575
Ford
Tempo
1989
473
700
A
78
Balt-Exet
B358
1G1AW51W8K6151308
Chevrolet
Celebri ty
1989
0
100
A
79
Balt-Exet
B361
JT2AE92W6J3076220
TOYATA
COROLLAWAGON
1988
A
80
Balt-Exet
B363
1G1JF14T2M713609 7
Chevrolet
Cavalier
1991
0
100
A
-------
OBS
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
11Z
113
114
115
116
117
118
119
120
Table B-l (Continued) 18:13 Friday, August 7, 1992 3
SITE
SAMPLE
VIN
HAKE
MODEL
YEAR
TPL
TPH
TF
Balt-Exet
B365
JT2AE83E4G3279918
TOYOTA
CAROLLA
1986
M
Balt-Exet
B367
1G2JB6903G7538623
PONT 1AC
SUNBIRD
1986
A
3alt-Exet
6368
KMHLF31J3JU42Q717
HYUNDAI
EXCEL
1988
A
3alt-Exet
B369
JE2AE72SXD2040392
TOYOTA
CAROLLA
1983
M
Balt-Exet
B370
JN1HJ01P7KT248360
Ni ssan
Maxima
1989
502
3977
A
3alt-Exet
B371
1N6ND0650EC332168
ni ssan
pickup
1984
A
3alt-Exet
B375
2G1AW35X3G1114186
Chevrolet
celebritygl
1986
A
3alt-Exet
B376
JN1HB1155CU006220
NISSAN
SENTRA
1982
M
3alt-Exet
B3B1
1G3H73735G1874913
OLDSMOBILE
88
1986
A
Balt-Exet
B386
JE2AE82E1G3308647
TOYOTA
COROLLA
1986
A
Balt-Exet
B389
JT2TE72W2C5109018
TOYOTA
COROLLA
1982
A
3alt-Exet
B392
2G4WB54T1M1805808
Bji ck
Regal
1991
100
A
Balt-Exet
B395
JN1PB1254FU631495
NISSAN
SENTRA
1985
M
Balt-Exet
B406
1G2NE27L1FC74B917
PONTIAC
GRANDAM
1985
A
Balt-Exet
B410
JM2UF1135J0369027
MAZDA
b20GO
1988
M
Balt-Exet
B413
4J47AAG110524
BUICK
REGAL
1980
A
Balt-Exet
8419
4J47WAG128222
BUICK
REGAL
1980
A
Balt-Exet
B420
1AB08C3DY262608
Chevrolet
Chevette
1983
A
Balt-Exet
B435
JB3CU14XXKU025841
Dodge
Colt
1989
317
2508
M
Balt-Exet
B436
4P3CS34TXHE109053
PLYMOUTH
LASER
1991
M
Balt-Exet
B438
JT2EL31GOH0069950
TOYOTA
TERCEL
1987
M
Balt-Exet
B441
1G1AB08C5EA125860
CHEVROLET
CHEVET
1984
A
Balt-Exet
B445
JT4YR29G5V5007140
TOYOTA
VAN
1986
M
Balt-Exet
B447
1G4AM69A7CH145805
BUICK
REGAL
1982
A
Balt-Exet
B451
1P3BP36K4HF153374
PLYMOUTH
RELIANTK
1987
A
3a 1 t-Exet
B454
2G4WD14W3K1400883
Buick
Regal
1989
100
A
Balt-Exet
B460
1FABP46F6D4168830
FORD
THUNDERBIRD
1983
A
Salt-Exet
B467
1HGBA743XGA063308
HONDA
ACCORD
1986
A
Balt-Exet
B468
JT2EL31D5H0075958
toyota
tercel
1987
M
Balt-Exet
B472
1P3BM18C6ED304919
PLYMOUTH
HORIZON
1984
M
Spokane
S002
1GCG<24M8GJ162254
CHEVROLET
SILVERAD020
1986
A
Spokane
S003
JN1HJ01P8LT366371
Nissan
Maxima
1990
533
3977
A
Spokane
S004
KMHLA31J8HU121264
hyundai
excelgs
1987
M
Spokane
S006
1G2AF54T9L6235376
Pontiac
6000 LE
1990
100
A
Spokane
S008
1FAPP36X7JK162901
ford
tempogl
1988
A
Spokane
son
1G3NT69U9GM372803
ol ds
Calais
1986
A
Spokane
S014
3G4AL54N3LS605881
Buick
Century
1990
0
100
A
Spokane
S015
HL41G8B367586
piymouth
volare
1978
A
Spokane
S018
GCFBAD246990
ford
fi esta
1980
M
Spokane
S019
JM1BG2265L 0111262
Mazda
Protege
1990
268
1937
A
-------
Table B-
1 (Continued)
OBS
SITE
SAMPLE
VIN
MAKE
MODEL
121
Spokane
S024
JF2KA83A2JD727208
SUBARU
JUSTY4WD
122
Spokane
S027
39128622
cadi 11 ac
coupedevi11e
123
Spokane
S028
1G2HX54C9L1241876
P0NTIAC
Bonneville
124
Spokane
$036
JT3FJ62G5J0087906
toyota
landcruiser
125
Spokane
S038
1G4AT27P2EK537437
buick
skylark
126
Spokane
S040
1G3AM54N2L6363462
01dsmobi1e
Cutlass
127
Spokane
S041
JT2AE92W1J3141485
toyota
corolla
128
Spokane
S045
1YVGD22B9L5241604
Mazda
626
129
Spokane
S046
2FTEF14N6GCA49786
ford
f 150
130
Spokane
S047
1B3XA463GLF914826
Dodge
Spi rit
131
Spokane
S048
AA0BE8S328524
Plymouth
traiIduster
132
Spokane
S050
1MEBP89C9EG646134
mercury
marqui s
133
Spokane
S051
JM1GD222XL1822669
Mazda
626
134
Spokane
S053
1FACP52U5LG167445
Ford
Taurus
135
Spokane
S054
1B7HW14T0GS016645
dodge
rampu
136
Spokane
S055
1J089AA196188
chevrol et
chevet
137
Spokane
S061
JM1BF222760145368
mazda
323
138
Spokane
S063
JT2SV21E2L0336391
Toyota
Camry
139
Spokane
S067
1C3BH58E1GN168729
chrysl er
labaron
140
Spokane
S069
JM1BF2327J0116975
mazda
323
141
Spokane
S070
JA3CR46V8LZ036160
Mitsubishi
Gal ant
142
Spokane
S072
2G2AK37H5E2268581
pontiac
grandpri x
143
Spokane
S073
JF2AN55B4GD439011
subaru
gl 10
144
Spokane
S078
1C3BA54E0EG237812
chrysl er
1 aser
145
Spokane
S088
JT2AE92E6J3034861
toyota
corolla
146
Spokane
S089
1B4FK44R0LX295520
Dodge
Caravan
147
Spokane
S090
1GNDL1526LB143279
Chevrolet
astro
148
Spokane
S092
1P3XA46K8LF815124
Plymouth
Acclaim
149
Spokane
S093
2G1WN54T4L9218037
Chevrolet
Lumi na
150
Spokane
S094
1G2WK14W3JF252210
pontiac
grandpri xle
151
Spokane
S097
2B4 FK453 0LR632 793
Dodge
Caravan
152
Spokane
S098
1G4CW54C8LI 624667
Buick
Park Avenue
153
Spokane
S100
1G1AW19R0G6112876
Chevrolet
eelebri ty
154
Spokane
S103
1G1LV14W9JE673479
Chevrolet
beretta
155
Spokane
S105
1FABP0521CW107600
ford
escort
156
Spokane
S107
1B3XA4639 LF838812
Dodge
Spi rit
157
Spokane
SI 09
JT2SV24E2L3411553
Toyota
Camry
158
Spokane
SI 12
1YVGD31BXL5201264
MAZDA
mx6
159
Spokane
SI 14
KMHLF21J9HU104162
hyundai
xl
160
Spokane
SI 17
JT2AE94A4L3346545
toyota
corolla
18:13 Friday, August 7, 1992
YEAR
TPL
TPH
TR
1988
1969
1990
1988
1984
1990
1988
1990
1986
1990
1978
1984
1990
1990
1986
1980
1986
1990
1986
1988
1990
1984
1986
1984
1988
1990
1990
1990
1990
1988
1990
1990
1986
1988
1982
1990
1990
1990
1987
1990
0
248
376
246
355
397
338
378
0
414
0
395
0
100
100
2181
1931
2185
2137
100
2487
1927
1852
100
1918
100
1904
100
M
A
A
A
A
A
H
A
M
A
A
A
M
A
A
M
A
A
A
M
A
A
M
A
A
A
A
A
A
A
A
A
A
A
M
A
A
A
M
A
-------
Table B-
1 (Continued)
OBS
SITE
SAMPLE
VIN
HAKE
MODEL
161
Spokane
SI IS
1G8CT1881E0141922
Chevrolet
blazer
162
Spokane
S119
1N4GB22B6LC764253
Ni ssan
Sentra
163
Spokane
S120
JT2SV22E0L3407469
toyota
carary
164
Spokane
S126
1Y15K5467 LZ153033
geo
prizm
165
Spokane
S128
1NXAE82G3J2543668
toyota
corolla
166
Spokane
S129
1YVGD22B6LS224064
Mazda
626
167
Spokane
S131
1MECH5344LG630501
Mercury
Sabl e
168
Spokane
S133
1G2NE5401LC373895
Pontiac
Grand Am
169
Spokane
S134
1HGCA554XJA105600
honda
accord
170
Spokane
S135
JB71M55E2JP049855
dodge
ram50
171
Spokane
S137
1G2PG9793GP263091
pontiac
fi ero
172
Spokane
SI 38
3N69FAR476562
oldsmobile
cutlasbrougham
173
Spokane
S143
JM1BF2321J0174161
MAZDA
323
174
Spokane
S144
1FA8P28A4DF128103
ford
mustang
175
Spokane
S147
JM2VF1133J0354042
rnazda
b2200pickup
176
Spokane
S148
2G1WL54T8L1197259
Chevrolet
Lumi na
177
Spokane
S156
1ZVPT20C1L5138646
ford
probe
178
Spokane
S158
1B3BA64E3EG128750
dodge
daytona
179
Spokane
S163
1GCEK14L5EJ110666
Chevrolet
clOpi ckup
180
Spokane
S164
1G3HY54C9LH309730
Oldsmobile
88
101
Spokane
S165
1N4GB22B9LC742425
ni ssan
sentra
182
Spokane
S167
1GCBS14C3G2140458
Chevrolet
slOpickup
183
Spokane
S169
JF1AX94241G316659
subaru
xt6
184
Spokane
S172
2B4FK41K2JR589231
dodge
caravan
185
Spokane
S175
JN1P52654EV633444
ni ssan
200sx
186
Spokane
SI 76
256690C110112
pontiac
executive
187
Spokane
S177
1G1LW14T8LE150508
Chevrolet
Beretta
188
Spokane
S178
JF1AC43B1JC230491
subaru
gi
189
Spokane
S179
1GCBS14E5G2149162
Chevrolet
slOpickup
190
Spokane
S180
JM2UC2217C0554694
mazda
b2000
191
Spokane
S181
3G4A454N7LS609966
Buick
Century
192
Spokane
S184
RA42061277
TOYOTA
CELICA
193
Spokane
S187
1GBCT18R5G8134828
Chevrolet
slOblazer
194
Spokane
S194
JM2UF211160523195
mazda
b2000
195
Spokane
S197
RN47029466
toyota
sr5pickup
196
Spokane
S200
X1X115A6114134
Chevrolet
citati or
197
Spokane
S201
1G3HY 54C9L181655 5
Oldsmobi le
88
198
Spokane
S203
1FABP3497GW162154
ford
escort
199
Spokane
S205
JM2UF1119G0634121
mazda
b2000
200
Spokane
S209
17A0987815
volkswagon
rabbit
18:13 Friday, August 7, 1992
YEAR
TPL
TPH
TR
1984
A
1990
456
4280
M
1990
A
1990
A
1986
M
1990
245
2175
A
1990
417
2026
A
1990
0
100
A
1988
H
1988
M
1986
A
1980
A
1988
M
1984
M
1988
M
1990
100
A
1990
A
1984
A
1984
A
1990
100
A
1990
M
1990
A
1988
M
1988
A
1984
M
1970
A
1990
100
M
1988
A
1986
M
1982
M
1990
100
A
1978
A
1986
M
1986
M
1980
M
1980
M
1990
0
100
A
1986
M
1986
M
1980
M
-------
OBS
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
Table B-l (Continued) 18:13 Friday, August 7, 1992 6
SITE
SAMPLE
VIN
MAKE
MODEL
YEAR
TPL
TPH
TF
Spokane
S211
JM1BF2228G0153432
mazda
323dx
1986
M
Spokane
S212
1M07VA7159960
chevrol et
monza
1980
A
Spokane
S214
1N6HD114XJC317900
ni ssan
pi ckup
1988
M
Spokane
S217
1G2WJ54T3LF263419
Ponti ac
Grand Prix
1990
0
100
A
Spokane
S218
SMK2089776
honda
accord
1980
M
Spokane
S219
1W27M8K407920
Chevrolet
mali buclassic
1978
A
Spokane
S221
1G1AD35P5EJ104025
chevrolet
cavali er
1984
A
Spokane
S222
JMV2F1136K0765114
mazda
b2200
1989
M
Spokane
S224
1FABP64T8JH140673
FORD
THUNDERBIRD
1988
M
Spokane
S227
1FABP40A4JF228320
ford
mustang! x
1988
M
Spokane
S229
JM2VC1218E0824500
mazda
b2000
1984
M
Spokane
S235
2G2AF51R0H9223691
ponti ac
6000
1987
A
Spokane
S23B
1G1LV14T1LE144679
Chevrol et
Beretta
1990
100
A
Spokane
S239
1G1JF77W1GJ208794
chevrolet
z24
1986
M
Spokane
S241
1FMDJ1541ELA53608
ford
bronco
1984
M
Spokane
S243
261WN54T9L9201394
Chevrolet
Lumina
1990
0
100
A
-------
-33 j
00 |
00 |
00 |
00 |
00 I
00 |
00 |
.00 I
.00 I
00 I
1
00
Table B-2 Speed & Acceleration Classes
TABLE OF SP5 BY AC1
18:13 Friday. August 7, 1992 8
-311 -281 -26| -?5| -22j -20| -19) -18) -16| -15| -H| -131 Total
4. 4. + \ + + + + + *- + +
0.00 I 0.C0 I 0.00 J 000 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 j O.OO J 0.00 I 0.00 I 19.35
4- h + 1- 4- +-- + + 4- 4- + - +
0.00 I 0.00 I 0.00 1 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 1 0.00 I 0.00 I 6.15
+ + + + + 4- + --- + + + + +
0.00 I 0.00 I O.OO I 0.00 1 0.00 1 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 5.50
+ + + + 4- + -f + + + — +
0.00 I 0.00 I 0.00 I 0.00 I O.OO I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 5.97
4. --+ + + +. + 4-— + + + I- +
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 6.17
1 + + 4 4. 4- + h + 1- +
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 j O.OO I 6.94
+ + 4. + + 4. + 4. 4. - +¦ + +
0.00 | 0.00 I 0.00 I 0.00 I 0.00 I 0.00 1 0.00 I 0.00 I 0.00 \ 0.00 I 0.00 I 0.00 I 9.18
+ + + + + + + + + + + 4
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 ) 0.00 I 0.00 I 0.00 I O.OO j 11.38
Y --4- + -I 4- + 1 + -*¦ + 4
0.00 I 0.00 I 0.00 I 0.00 I 0.00 j 0.00 I 0.00 I 0.00 I 0 00 I 0.00 I 0.00 I 0.00 I 8.50
+ + ^- + + 4- 4- --4 + 4- + +
0.00 I 0.00 [ 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 i 0.00 ( 0.00 [ O.OO I 0.00 [ 4.56
4. 4- 4. + + 4- 4- + + H + +
1 1 1 1 1 1 1 1 1 2 S 16 6915785
0.00 0.00 0.00 0 00 0.00 0.00 0.00 0.00 0.00 0 00 0.00 0.00 100.DO
-------
Table B-2 (Continued) Speed & Acceleration Classes
TABLE OF SP5 BY AC1
18:13 Friday, August 7, 1992 9
SP5 AC1
Percent | -331 -311 -281 -261 -251 -221 -20| -191 -18f -16| -15| -141 -131 Total
50 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 [ 0.00 | 0.00 | 2.96
55 I 0.00 I 0.00 I 0.00 I 0.00 | 0.00 | 0.00 | 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0,00 I 2.95
60 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 4.21
65 I 0.00 I O.OO I 0.00 I 0.00 | 0.00 | 0.00 | 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 3.51
—+ + + + + + + + + + + + + +
70 I 0.00 I 0.00 I 0.00 j 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 j 0.00 I 1.85
75 I 0.00 I 0.00 I 0.00 I 0.00 | 0.00 | 0.00 | 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.66
80 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.14
+ + + + + + + + + + + + + +
85 I 0.00 j 0.00 I 0.00 I 0.00 | 0.00 | 0.00 | 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.02
90 I 0.00 I 0.00 I 0.00 I 0.00 I 0.DO | 0.00 | 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00
95 I 0.00 I 0.00 I 0.00 I 0.00 | 0.00 | 0.00 | 0.00 I 0.00 I 0.00 I 0.00 I O.OO I 0.00 I 0.00 I 0.00
Total 1 1 1 1 1 1 1 1 1 1 2 8 16 6915785
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
(Continued)
-------
Table B-2 (Continued) Speed & Acceleration Classes
TABLE OF SP5 BY AC1
18:13 Friday, August 7, 1992 10
SP5 AC1
Percent |
"121
-111
-10|
-9|
"81
-A
-6|
"5|
-4|
"3i
"2|
-1|
o|
Total
0 1
0.00 I
0.00 1
0.00 1
0.00 1
0.00 1
0.00 1
0.01 |
0.02 |
0.04 |
0.12 |
0.23 |
0.39 |
18.54 |
19.35
5 1
0.00 I
0.00 I
0.00 1
0.00 I
O.OO 1
0.00 1
0.02 |
0 05 |
0.12 |
0.31 |
0.54 |
0.86 |
1.44 |
6.15
10 |
0.00 1
0.00 1
0.00 I
0.00 1
0.00 1
0.02 |
0.04 |
0.12 |
0.24 |
0.40 |
0.50 |
0.64 |
1.04 |
5.50
15 |
0.00 I
0.00 1
0.00 1
0.00 1
0.01 1
0.02 1
0.05 |
0.13 |
0.27 |
0.43 |
0.48 |
0.59 |
0.98 |
5.97
20 |
0.00 I
O.OO 1
0.00 1
0.00 1
0.00 I
0.01 I
0.03 |
0.11 |
0.24 |
0.41 |
0.48 |
0.59 |
1.00 |
6.17
25 |
0.00 I
0.00 I
0.00 I
0.00 I
0.00 j
0.01 I
0.02 |
0.06 |
0.16 |
0.33 j
0.45 |
0.70 |
1.52 |
6.94
30 |
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.01 |
0.03 |
0.08 |
0.21 |
0.37 |
0.87 |
2.99 |
9.18
35 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 j
0.00 I
0.00 |
0.01 |
0.03 |
0.10 |
0.21 j
0.81 |
4.72 |
11.38
to |
0.00 I
0.00 1
0.00 1
0.00 I
0.00 I
0.00 I
0.00 |
0.01 |
0.01 |
0.06 |
0.10 |
0.51 |
3.68 |
8.50
45 |
0.00 j
0.00 1
0.00 1
0.00 1
O.OO I
0.00 1
0.00 |
0.00 |
0.01 |
0.02 |
0.05 j
0.29 |
1.95 |
4.56
Total
40
0.00
85
0.00
174
0.00
449
0.01
1569
0.02
4578
0.07
12959
0.19
374B9
0.54
B4196
1.22
167796
2.43
241242
3.49
494090
7.14
3180831 6915785
45.99 100.00
(Continued)
-------
-12]
00 |
.00 I
.00 I
.00 I
.00 I
.00 I
.00 I
.00 I
.00 I
00 I
40
00
Table B-2 (Continued) Speed & Acceleration Classes 18:13 Friday, August 7, 1992 11
TABLE OF SP5 BY AC1
-ll| -101 -91 -8| -71 -6| -5| -41 -31 -2| -l| 0| Total
0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.01 | 0.03 j 0.19 | 1.30 | 2.96
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.01 I 0.02 I 0.17 I 1.41 | 2.95
+ + + + + + + + + + + +
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.01 I 0.01 I 0.18 I 2.18 | 4.21
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 i 0.00 I 0.00 I 0.17 I 1.87 I 3.51
+ + + + + + + + + + + +
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.10 I 0.98 I 1.85
+ + + + + + + + + + + +
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.07 I 0.33 I 0.66
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.02 I 0.07 I 0.14
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.01 I 0.02
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00
0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00
+ + + + + + + + + + + +
85 174 449 1569 4578 12959 37489 84196 167796 241242 494090 3180831 6915785
0.00 0.00 0.01 0.02 0.07 0.19 0.54 1.22 2.43 3.49 7.14 45.99 100.00
-------
Table B-2 (Continued) Speed & Acceleration Classes
TABLE OF SP5 Br AC1
18:13 Friday, August 7. 1992 12
SP5 AC1
Percent | l| 2\ 3| 4| 5| 6| 7| B| 9| 10( 11| 12| 13| Total
0 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 I 0.00 I O.OO j 0-00 | 0.00 | 0 00 | 0.00 | 0.00 | 19.35
5 | 1.73 | 0.53 | 0.29 | 0.18 | 0.07 | O.OO | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 6.15
10 | 0.82 j 0.53 | 0.39 j 0.30 | 0.19 | 0.13 | 0.06 | 0.03 j 0.02 j O.Oi | 0.00 | 0.00 | 0.00 | 5.50
15 | 0.94 | 0.76 | 0.58 | 0.40 | 0.20 | 0.09 | 0.03 | 0.01 | 0.00 | 0.00 | 0.01 | 0.00 | 0.00 | 5.97
20 | 1.01 | 0.93 | 0.75 | 0.41 | 0.13 | 0.04 | 0.02 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 6.17
25 | 1.49 | 1.03 | 0.69 | 0.34 | 0.09 | 0.03 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 6.94
30 | 2.75 | 1.15 | 0.46 | 0.19 j 0.05 | 0.02 | 0.01 | 0.00 | 0.00 | 0.00 | 0.00 ] 0.00 [ 0.00 | 9.18
35 j 4.19 | 0.94 | 0.26 | 0.07 | 0.01 | 0.01 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 11.38
40 | 3.40 | 0.56 | 0.11 | 0.05 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 j 0.00 | 8.50
45 | 1.87 | 0.30 | 0.05 | 0.02 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 [ 0.00 | O.OO | 0.00 | 4.56
Total 1698887 514091 251894 136891 51915 22347 7907 3499 1556 628 366 136 57 6915785
24.57 7.43 3.64 1.98 0.75 0.32 0.11 0.05 0.02 0.01 0.01 0.00 0.00 100.00
(Continued)
-------
Table B-2 (Continued) Speed & Acceleration Classes 18:13 Friday, August 7, 1992 13
TABLE OF
SP5 BY
AC1
SP5
AC1
Percent
1 1|
2|
3|
<1
5|
6|
7|
8|
9|
10|
111
12 I
13 I
Total
50
1 1-23 |
0.17 |
0.02 |
0.01 |
0.00 I
0.00 I
0.00 I
0.00 1
0.00 I
0.00 1
0.00 I
0.00 I
0.00 1
2.96
55
1 1-20 |
0.13 |
0.01 |
0.01 |
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
2.95
60
1 1.71 |
0.11 |
0.01 |
0.00 |
0.00 I
0.00 1
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 1
0.00 I
4.21
65
| 1.33 |
0.11 |
0.00 |
0.00 1
0.00 I
0.00 I
0.00 |
0.00 1
0.00 I
0.00 I
0.00 1
0.00 1
0.00 1
3.51
70
| 0.68 |
0.08 |
0.00 |
0.00 1
0.00 1
0.00 I
0.00 I
0.00 I
0.00 1
0.00 1
0.00 I
0.00 1
0.00 i
1.85
75
| 0.18 |
0.07 |
0.00 I
0.00 I
0.00 I
0.00 1
0.00 I
0.00 I
0.00 1
0.00 1
O.OO 1
0.00 1
0.00 1
0.66
80
85
| 0.03 |
+ + •
| 0.00 I
0.03 |
0.00 |
0.00 I
+-
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
+-
0.00 I
0.00 1
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
O.OO I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.14
0.02
90
95
| 0.00 I
-+ +-
| 0.00 I
0.00 |
0.00 |
0.00 I
0.00 I
0.00 I
0.00 1
0.00 I
0.00 1
0.00 I
0.00 I
0.00 I
+-
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00
0.00
Total 1698887
24.57
(Continued)
514091
7.43
251894
3.64
136891
1.98
51915
0.75
22347
0.32
7907
0.11
3499
0.05
1556
0.02
628
0.01
366
0.01
136
0.00
57 6915785
0.00 100.00
-------
Table B-2 (Continued) Speed & Acceleration Classes
TABLE OF SP5 BY AC1
18:13 Friday, August 7, 1992 14
SP5 AC1
Percent | 14) 15| 1B| 17| 13] 19| 20) 261 27| 31J 32J 331 34j Total
4 + + + + - + 4 4 + H 4 4 + 4
0 j 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 ! 0 00 [ 19.35
+ 4 4 + + + + H + + 4 + + +
5 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 6.15
10 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I O.OO I 0.00 I 0.00 I 0.00 I 5.50
+ + + + + 4 + + + + + 4 + +
15 I 0.00 I 0.00 ] 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 5.97
+ 4. + + 4- + + + + 4. 4 + + +
20 I 0.00 j 0.00 | 0.00 I 0.00 j 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 6.17
+ + + + 4 + 4 + 4 + 4 4 4 +
25 I 0.00 I 0.00 I 0.00 I 0.00 j 0.00 | 0.00 | 0.00 I O.OO I 0.00 I 0.00 ( 0.00 I 0.00 I 0.00 I 6.94
+. + 4 4 + + + + 4 + + + + +
30 I 0.00 I 0.00 I 0.00 j 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 i 0 00 I 9.18
H 4 4 + h 4 4- 4- ^ + 4 4 4 4-
35 I 0.00 I 0.00 I 0.00 I 0.00 | 0.00 | 0.00 | 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 11.38
+ + 4 4 4 4 4 4 4 4 4 4 4 +
40 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0 00 I 8.50
4 4 4 4 4 4 4 4 4 4 4 4 4 4
45 I 0.00 I 0.00 I 0.00 I 0.00 | 0.00 | 0.00 | 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 0.00 I 4.56
¦" — — — — — —4~-~ — ~^ — — — — + - — — — — — — — 4— — - - _ _ __ 4_-».. _ _ _ __ — _ 4 H— — — — — — — — 4— - — — — — 4- — — — — +
Total 26 16 12 5 6 2 4 1 1 1 1 1 1 6915785
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
(Conti nued)
-------
Table B-2 (Continued) Speed & Acceleration Classes
TABLE OF SP5 BY AC1
18:13 Friday, August 7, 1992 15
SP5
AC1
Percent
1 141
15|
161
17|
181
191
201
26 (
271
31 I
32 I
331
341
Total
50
| 0.00 |
0.00 I
0.00 I
0.00 I
0.00 I
0.00 1
0.00 1
0.00 1
0.00 1
0.00 I
0.00 1
0.00 1
0.00 I
2.96
55
| 0.00 1
0.00 1
0.00 1
0.00 1
0.00 1
0.00 1
0.00 1
0.00 1
0.00 1
0.00 1
0.00 1
0.00 1
0.00 1
2.95
60
I 0.00 1
0.00 1
0.00 I
0.00 1
0.00 I
0.00 I
0.00 I
0.00 I
0.00 1
0.00 I
0.00 1
0.00 I
0.00 I
4.21
65
| 0.00 I
0.00 I
0.00 I
0.00 1
0.00 1
0.00 i
0.00 1
0.00 1
0.00 1
0.00 I
0.00 1
0.00 1
0.00 I
3.51
70
| 0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 1
0.00 I
0.00 I
1.85
75
| 0.00 I
0.00 I
0.00 I
0.00 I
0.00 1
0.00 I
0.00 I
0.00 1
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.66
80
| 0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 1
0.00 I
0.00 I
0.14
85
| 0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.02
90
| 0.00 1
0.00 I
0.00 I
0.00 1
0.00 1
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 1
0.00 I
0.00 I
0.00
95
| 0.00 I
0.00 I
0.00 I
0.00 I
0.00 1
0.00 I
0.00 I
0.00 I
0.00 I
0.00 I
0.00 1
0.00 I
0.00 I
0.00
Total
26
0.00
16
0.00
12
0.00
5
0.00
6
0.00
2
0.00
4
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1 6915785
0.00 100.00
Frequency Missing = 24626
-------
Table B-3
Speed Frequency
Di stri buti on
IBS
LOWER
UPPER
COUNT
PERCENT
CUM
DECUM
1
0
0
1348267
19.4634
19.46
80.54
2
0
5
426249
6.1533
25.61
74.39
3
5
10
380510
5.4930
31.10
68.90
4
10
15
412799
5.9591
37.06
62.94
5
15
20
426868
6.1622
43.22
56.78
6
20
25
479650
6.9241
50.14
49.86
7
25
30
635168
9.1692
59.31
40.69
8
30
35
787139
11.3630
70.67
29.33
9
35
40
588185
8.4909
79.16
20.84
10
40
45
315393
4.5530
83.71
16.29
11
45
50
204782
2.9562
86.67
13.33
12
50
55
204206
2.9479
89.62
10.38
13
55
60
291118
4.2025
93.82
6.18
14
60
65
242497
3.5006
97.32
2.68
15
65
70
127742
1.8441
99.16
0.84
16
70
75
45645
0.6589
99.82
0.18
17
75
80
9674
0.1397
99.96
0.04
18
80
85
1175
0.0170
99.98
0.02
19
85
90
128
0.0018
99.98
0.02
20
90
95
11
0.0002
99.98
0.02
18:13 Friday, August 7. 1992 16
-------
Table B-4 Acceleration Frequency Distribution
OBS LOWER UPPER
1
-34
-33
2
-32
-31
3
-29
-28
4
no
-26
5
-26
-25
6
-23
-22
7
-21
-20
a
-20
-19
9
-19
-18
ID
-17
-16
11
-16
-15
12
-15
-14
13
-14
-13
14
-13
-12
15
-12
-11
16
-11
-10
17
-10
-9
18
-9
-8
19
-8
-7
20
-7
-6
21
-6
-5
22
-5
-4
23
-4
-3
24
-3
-2
25
-2
-1
26
-1
0
27
0
1
28
1
2
29
2
3
30
3
4
31
4
5
32
5
6
33
6
7
34
7
8
35
8
9
36
9
10
37
10
11
38
11
12
39
12
13
40
13
14
PERCENT CUM
0.0000 O.OO
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.0D
0.0000 0.00
0.0000 0.00
0.0000 0.00
0 0000 0.00
0.0000 0.00
0.0001 0.00
0.0002 0.00
0.0006 0.00
0.0012 0.00
0.0025 0.00
0.0065 0.01
0.0227 0.03
0.0662 0.10
0.1874 0.29
0.5421 0.83
1.2174 2.05
2.4263 4.48
3.4883 7.97
7.1444 15.11
45.9938 61.10
24.5654 85.67
7.4336 93.10
3.6423 96.74
1.9794 98.72
0.7507 99.47
0.3231 99.79
0.1143 99.90
0.0506 99.95
0.0225 99.97
0.0091 99.98
0.0053 99.99
0.0020 99.99
0.0008 99.99
0.0004 99.99
COUNT
1
1
1
1
1
1
1
1
1
1
2
8
16
40
85
174
449
1569
45/8
12959
37489
84196
167796
241242
494090
3180831
1698887
514091
251894
136891
51915
22347
7907
3499
1556
628
365
136
57
26
18:13 Friday, August 7, 1992
DECUM
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
99.99
99.97
99.90
99.71
99.17
97.95
95.52
92.03
84.89
38.90
14.33
6.90
3.26
1.28
0.53
0.21
0.10
0.05
0.03
0.02
0,01
0.01
0 01
0.01
-------
Table B-4 (Continued)
OBS LOWER
41
14
42
15
43
16
44
17
45
18
46
19
47
25
48
26
49
30
50
31
51
32
52
33
on Frequency Distribution
COUNT PERCENT CUM
16
.00023135
99.99
12
.00017352
99.99
5
.00007230
99.99
6
.00008676
99.99
2
.00002892
99.99
4
.00005784
99.99
1
.00001446
99.99
1
.00001446
99.99
1
.00001446
99.99
1
.00001446
99.99
1
.00001446
99.99
1
.00001446
99.99
Accelerati
UPPER
15
16
17
18
19
20
26
27
31
32
33
34
IB:13 Friday, Ajgust 7, 1992
CECUM
0.01
0.01
0.01
0.01
0.01
0.01
0-01
0.01
0.01
0.01
0.01
0.01
-------
Table B-5
Power
Frequency
OBS
LOWER
UPPER
COUNT
1
0
20
914113
2
20
40
549123
3
40
60
400522
4
60
80
291349
5
80
100
190744
6
100
120
125479
7
120
140
80238
8
140
160
48698
9
160
180
27911
10
180
200
17295
11
200
220
14588
12
220
240
13465
13
240
260
6413
14
260
280
4461
15
280
300
1954
16
300
320
1033
17
320
340
850
18
340
360
588
19
360
380
393
20
380
400
307
21
400
420
156
22
420
440
104
23
440
460
176
24
460
480
114
25
480
500
56
26
500
520
48
27
520
540
22
28
540
560
15
29
560
580
"6
30
580
600
10
31
600
620
7
32
620
640
4
33
640
660
1
34
660
680
1
35
700
720
1
36
880
900
1
37
1420
1440
1
38
1580
1600
1
39
2700
2720
1
40
2720
2740
1
i str i but ion
18:13 Friday, August 7, 1992 20
PERCENT
33.9787
20.4116
14.8879
10.8298
7.0902
4.6642
2.9825
1.8102
1.0375
0.6429
0.5423
0.5005
0.2384
0.1658
0.0726
0.0384
0.0316
0.0219
0.0146
0.0114
0.0058
0.0039
0.0065
0.0042
0.0021
0.0018
0.0008
0.0006
0.0002
0.0004
0.0003
0.0001
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
CUM DECUM
33.98 66.02
54.39 45.61
69.28 30.72
80.11 19.89
87.20 12.80
91.86 8.14
94.84 5.16
96.65 3.35
97.69 2.31
98.33 1.67
98.87 1.13
99.37 0.63
99.61 0.39
99.78 0.22
99.85 0.15
99.89 0.11
99.92 0.08
99.94 0.06
99.95 0.05
99.96 0.04
99.97 0.03
99.97 0.03
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
99.98 0.02
-------
Table B-5 (Continued) Power Frequency Distribution
OBS LOWER UPPER COUNT PERCENT CUM OECUM
41 2860 2880 1 .000037171 99.98 0.02
18:13 Friday, August 7, 1992 21
-------
Figure B-6
PERCENT
01 2 J 5 7 8 9 1 12233445567881 1
--------050305050500000 1 I
I 23468ft 1 - -- -- -- -- -- - 002
0 1 223344556789 1 - -0
30505050500000 1 1
0 1 2
0 0
Trip Time (minutes)
-------
Table B-7 Distribution of Trip Time
Cumulative Cumulative Decumulative
CTTLIN Frequency Percent Frequency Percent Percent
15.6 1879 15.6% 84.4%
6.3 2645 21.9% 78.1%
6.3 3409 28.2% 71.8%
6.6 4203 34.8% 65.2%
6.4 4972 41.2% 58.8%
5.4 5620 46.6% 53.4%
5.3 6254 51.8% 48.2%
4.6 6804 56.4% 43.6%
4.1 7301 60.5% 39.5%
4.3 7820 64.8% 35.2%
14.4 9562 79.2% 20.8%
8.7 10618 87.9% 12.1%
5 11220 92.9% 7.1%
2.8 11559 95.7% 4.3%
1.6 11751 97.3% 2.7%
1 11867 98.3% 1.7%
0.5 11925 98.8% 1.2%
0.3 11965 99.1% 0.9%
0.2 11995 99.4% 0.6%
0.2 12019 99.6% 0.4%
0.1 12037 99.7% 0.3%
0.1 12051 99.8% 0.2%
0.1 12058 99.9% 0.1%
0 12061 99.9% 0.1%
0 12063 99.9% 0.1%
0 12067 100.0% 0.0%
0 12073 100.0% 0.0%
0-1
1879
1-2
766
2-3
764
3-4
794
4-5
769
5-6
648
6-7
634
7-8
550
8-9
497
9-10
519
10-15
1742
15-20
1056
20-25
602
25-30
339
30-35
192
35-40
116
40-45
58
45-50
40
50-55
30
55-60
24
60-70
18
70-80
14
80-90
7
90-100
3
100-110
2
110-120
4
>120
6
-------
Table B-8
Trip Time
CI asses
(Minutes)
OBS LOWER
1
0.200
2
0.252
3
0.317
4
0.399
5
0.502
6
0.632
7
0.796
8
1.002
9
1.262
10
1.589
11
2.000
12
2.518
13
3.170
14
3.991
15
5.024
16
6.325
17
7.962
18
10.024
19
12.619
20
15.887
21
20.000
22
25.179
23
31.698
24
39.905
25
50.238
26
63.246
27
79.621
28
100.237
29
126.191
30
158.866
UPPER COUNT
0.252 1230
0.317 61
0.399 58
0.502 113
0.632 114
0.796 136
1.002 167
1.262 191
1.589 239
2.000 336
2.518 387
3.170 525
3.991 635
5.024 786
6.325 853
7.962 951
10.024 1044
12.619 1036
15.887 929
20.000 827
25.179 613
31.698 405
39.905 230
50.238 101
63.246 58
79.621 26
100.237 10
126.191 7
158.866 3
200.000 2
PERCENT CUM
10.1880 10 19
0.5053 10.70
0.4804 11.18
0.9360 12.12
0.9443 13.06
1.1265 14.19
1.3833 15.57
1.5820 17.15
1.9796 19.13
2.7831 21.91
3.2055 25.12
4.3485 29.47
5.2597 34.73
6.5104 41.24
7.0654 48.31
7.8771 56.19
8.6474 64.84
8.5811 73.42
7.6949 81.11
6.8500 87.96
5.0774 93.04
3.3546 96.39
1.9051 98.30
0.8366 99.14
0.4804 99.62
0.2154 99.84
0.0828 99.92
0.0580 99.98
0.0248 100.00
0.0166 100.02
00:15 Tuesday, August 11, 1992 1114
DECUM
89.81
89.30
88.82
87.88
86.94
85.81
84.43
82.85
80.87
78.09
74.88
70.53
65.27
58.76
51.69
43.81
35.16
26.58
18.89
12.04
6.96
3.61
1.70
0.86
0.38
0.16
0.08
0.02
0.00
-0.02
-------
Figure B-9
PERCENT
40
30
20
10
0
0
1
2
3
4
5
6
7
a
•
1
2
3
4
a
6
7
a
9
>
_
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
—
—
—
0
0
0
0
0
0
0
0
0
0
0
2
3
4
3
•
7
8
9
1
—
—
—
-
-
-
-
-
-
0
0
0
0
0
0
0
0
0
0
2
3
4
3
6
7
a
0
t
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Soak Time (minutes)
-------
Table B-10 Distribution of Soak Time
Cumulat ive
Cumulative
Decumulati
CSTLIN
F requency
Percent
Frequency
Percent
Percent
0-10
4294
35.6
4294
36.2%
63.8%
10-20
1093
9.1
5387
45.4%
54.6%
20-30
677
5.6
6064
51.1%
48.9%
30-40
518
4.3
6582
55.5%
44.5%
40-50
385
3.2
6967
58.8%
41.2%
50-60
321
2.7
7288
61.5%
38.5%
60-70
259
2.1
7547
63.7%
36.3%
70-80
224
1.9
7771
65.5%
34.5%
80-90
189
1.6
7960
67.1%
32.9%
90-100
159
1.3
8119
68.5%
31.5%
100-200
1033
8.6
9152
77.2%
22.8%
200-300
484
4
9636
81.3%
18.7%
300-400
207
1.7
9843
83.0%
17.0%
400-500
212
1.8
10055
84.8%
15.2%
500-600
364
3
10419
87.9%
12.1%
600-700
228
1.9
10647
89.8%
10.2%
700-800
252
2.1
10899
91.9%
8.1%
800-900
256
2.1
11155
94.1%
5.9%
900-1000
195
1.6
11350
95.7%
4.3%
>1000
507
4.2
11857
100.0%
0.0%
-------
Table B-ll Soak time (Minutes) Classes
OBS LOWER
1
0.10
2
0.16
3
0.25
4
0.40
5
0.63
6
1.00
7
1.59
8
2.51
9
3.98
10
6.31
11
10.00
12
15.85
13
25.12
14
39.81
15
63.10
16
100.00
17
15B.49
18
251.19
19
398.11
20
630.96
21
1000.00
22
1584.89
23
2511.89
24
3981.07
25
6309.57
UPPER COUNT
0.16 708
0.25 47
0.40 570
0.63 58
1.00 101
1.59 231
2.51 440
3.98 683
6.31 694
10.00 762
15.85 724
25.12 769
39.81 780
63.10 801
100.00 751
158.49 702
251.19 609
398.11 410
630.96 659
1000.00 851
1584.89 393
2511.89 56
3981.07 44
6309.57 13
10000.00 1
PERCENT CUM
5.97116 5.97
0.39639 6.37
4.80729 11.18
0.48916 11.67
0.85182 12.52
1.94822 14.47
3.71089 18.18
5.76031 23.94
5.85308 29.79
6.42658 36.22
6.10610 42.33
6.48562 48.82
6.57839 55.40
6.75550 62.16
6.33381 68.49
5.92055 74.41
5.13621 79.55
3.45787 83.01
5.55790 88.57
7.17719 95.75
3.31450 99.06
0.47229 99.53
0.37109 99.90
0.10964 100.01
0.00843 100.02
00:15 Tuesday, August 11, 1992 1115
DECUM
94.03
93.63
88.82
88.33
87.48
85.53
81.82
76.06
70.21
63.78
57.67
51.18
44.60
37.84
31.51
25.59
20.45
16.99
11.43
4.25
0.94
0.47
0.10
-0.01
-0.02
-------
Figure B-12
PERCENT
60
50
40
30
20
10
0
0
1
2
3
3
7
a
9
1
1
2
2
3
J
4
4
5
S
6
7
8
9
t
1
>
—
—
—
—
—
—
—
—
0
a
0
5
0
5
0
5
0
5
0
0
0
0
0
t
\
1
2
3
4
e
e
0
t
—
-
—
—
—
—
—
—
—
—
-
—
-
0
0
2
0
1
2
2
3
3
4
4
S
s
6
7
8
9
t
-
-
0
5
0
8
0
5
0
5
0
5
0
0
0
0
0
0
1
t
0
1
2
0
Time in Idle (minutes)
-------
Table B-13 Distribution of Idle Time
Cumulative
CJTLIN Frequency Percent Frequency
0
47
0.4
47
0-1
5741
47.6
5788
1-2
2361
19.6
8149
2-3
1246
10.3
9395
3-4
692
5.7
10087
4-5
425
3.5
10512
5-6
291
2.4
10B03
6-7
176
1.5
10979
7-8
122
1
11101
8-9
98
0.8
11199
9-10
54
0.4
11253
10-15
192
1.6
11445
15-20
52
0.4
11497
20-25
14
0.1
11511
25-30
7
0.1
11518
30-35
2
0
11520
35-40
5
0
11525
40-45
1
0
11526
45-50
1
0
11527
50-55
1
0
11528
>55
4
0
11532
Cumulative Decinulative
Percent Percent
0.4%
99.6%
50.2%
49.8%
70.7%
29.3%
81.5%
18.5%
87.5%
12.5%
91.2%
8.8%
93.7%
6.3%
95.2%
4.8%
96.3%
3.7%
97.1%
2.9%
97.6%
2.4%
99.2%
0.8%
99.7%
0.3%
99.8%
0.2%
99.9%
0.1%
99.9%
0.1%
99.9%
0.1%
99.9%
0.1%
100.0%
0.0%
100.0%
0.0%
100.0%
0.0%
-------
Table B-14 Time in Idle (Mi
OBS LOWER
1
0.040
2
0.050
3
0.063
4
0.080
5
0.100
6
0.126
7
0.159
8
0.200
9
0.252
10
0.318
11
0.400
12
0.504
13
0.634
14
0.798
15
1.005
16
1.265
17
1.592
18
2.005
19
2.524
20
3.177
21
4.000
22
5.036
23
6.340
24
7.981
25
10.048
26
12.649
27
15.924
28
20.047
29
25.238
30
31.773
UPPER COUNT
0.050 848
0.063
0.080 116
0.100 402
0.126 145
0.159 255
0.200 374
0.252 314
0.318 407
0.400 466
0.504 477
0.634 579
0.798 606
1.005 754
1.265 790
1.592 751
2.005 820
2.524 749
3.177 651
4.000 539
5.036 434
6.340 346
7.981 234
10.048 156
12.649 130
15.924 69
20.047 42
25.238 14
31.773 7
40.000 10
es) Classes
00:15 Tuesday, August 11, 1992 1116
PERCENT
7.38354
0.00000
1.01001
3.50022
1.26252
2.22029
3.25642
2.73400
3.54375
4.05747
4.15324
5.04136
5.27645
6.56508
6.87854
6.53896
7.13975
6.52155
5.66826
4.69308
3.77884
3.01263
2.03744
1.35829
1.13191
0.60078
0.36569
0.12190
0.06095
0.08707
CUM DECUH
7.38
92.62
7.38
92.62
8.39
91.61
11.89
88.11
13.15
86.85
15.37
84.63
18.63
81.37
21.36
78.64
24.90
75.10
28.96
71.04
33.11
66.89
38.15
61.85
43.43
56.57
50.00
50.00
56.88
43.12
63.42
36.58
70.56
29.44
77.08
22.92
82.75
17.25
87.44
12.56
91.22
8.78
94.23
5.77
96.27
3.73
97.63
2.37
98.76
1.24
99.36
0.64
99.73
0.27
99.85
0.15
99.91
0.09
100.00
0.00
-------
Figure B-15
PERCENT
17
Running Time (minutes)
-------
Table B-16 Distribution of Running Time
Cumulative Cumulative Decumulative
CRTLIM Frequency Percent Frequency Percent Percent
0
715
5.9
715
6.2%
93.8%
0-1
983
8.1
1698
14.7%
85.3%
1-2
1004
8.3
2702
23.4*
76.6%
2-3
917
7.6
3619
31.4*
68.6%
3-4
945
7.8
4564
39.6%
60.4%
4-5
775
6.4
5339
46.3%
53.7%
5-6
750
6.2
6089
52.8%
47.2%
6-7
643
5.3
6732
58.4%
41.6%
7-8
560
4.6
7292
63.3%
36.A
8-9
551
4.6
7843
68.0%
32.0%
9-10
447
3.7
8290
71.9%
28.1%
10-15
1551
12.9
9841
85.4%
14.6%
15-20
784
6.5
10625
92.2%
7.8%
20-25
423
3.5
11048
95.8%
4.2%
25-30
203
1.7
11251
97.6%
2.4%
30-35
96
0.B
11347
98.4%
1.6%
35-40
52
0.4
11399
98.9%
1.1%
40-45
34
0.3
11433
99.2%
0.8%
45-50
24
0.2
11457
99.4%
0.6%
50-55
26
0.2
11483
99.6%
0.4%
55-60
7
0.1
11490
99.7%
0.3%
60-70
18
0.1
11508
99.8%
0.2%
70-80
9
0.1
11517
99.9%
0.1%
80-90
3
0
11520
99.9%
0.1%
90-100
1
0
11521
99.9%
0.1%
100-110
5
0
11526
100.0%
0.0%
110-120
2
0
11528
100.0%
0.0%
-------
Table B-17 Time in Running (Hi
OBS
LOWER
UPPER
COUNT
1
0.200
0.252
329
2
0.252
0.317
71
3
0.317
0.399
70
4
0.399
0.502
122
5
0.502
0.632
109
6
0.632
0.796
125
7
0.796
1.002
157
8
1.002
1.262
240
9
1.262
1.589
339
10
1.589
2.000
425
11
2.000
2.518
481
12
2.518
3.170
585
13
3.170
3.991
777
14
3.991
5.024
807
15
5.024
6.325
951
16
6.325
7.962
963
17
7.962
10.024
1030
18
10.024
12.619
932
19
12.619
15.8B7
789
20
15.887
20.000
608
21
20.000
25.179
432
22
25.179
31.698
239
23
31.698
39.905
102
24
39.905
50.238
60
25
50.238
63.246
39
26
63.246
79.621
20
27
79.621
100.237
4
28
100.237
126.191
7
29
126.191
158.866
2
30
158.866
200.000
2
) Classes
00:15 Tuesday, August 11, 1992 1117
PERCENT
3.04151
0.65637
0.64713
1.12785
1.00767
1.15559
1.45142
2.21873
3.13396
3.92900
4.44670
5.40815
7.16314
7.46048
8.79172
8.90265
9.52205
8.61607
7.29407
5.62078
3.99371
2.20949
0.94296
0.55468
0.36054
0.18489
0.03698
0.06471
0.01849
0.01849
CUM DECUM
3.04
96.96
3.70
96.30
4.35
95.65
5.48
94.52
6.49
93.51
7.65
92.35
9.10
90.90
11.32
88.68
14.45
85.55
18.38
81.62
22.83
77.17
28.24
71.76
35.42
64.58
42.88
57.12
51.67
48.33
60.57
39.43
70.09
29.91
78.71
21.29
86.00
14.00
91.62
8.38
95.61
4.39
97.82
2.18
98.76
1.24
99.31
0.69
99.67
0.33
99.85
0.15
99.89
0.11
99.95
0.05
99.97
0.03
99.99
0.01
-------
Figure B-18
Trip Distance (miles)
-------
Table B-19 Distribution of Trip Distance
Cumulative
Cumulativi
CDSLIN
Frequency
Percent
F requency
Percent
0
1256
10.4
1256
10.4%
0-0.1
794
6.6
2050
17.0%
0.1-0.2
246
2
2296
19.0%
0.2-0.3
277
2.3
2573
21.3%
0.3-0.4
351
2.9
2924
24.2%
0.4-0.5
253
2.1
3177
26.3%
0.5-0.6
258
2.1
3435
28.5%
0.6-0.7
239
2
3674
30.4%
0.7-0.8
229
1.9
3903
32.3%
0.8-0.9
246
2
4149
34.4%
0.9-1
184
1.5
4333
35.9%
1-1.2
397
3.3
4730
39.2X
1.2-1.4
378
3.1
5108
42.3%
1.4-1.6
379
3.1
5487
45.4%
1.6-1.8
397
3.3
5884
48.7%
1.8-2
325
2.7
6209
51.4%
2-3
1296
10.7
7505
62.2%
3-4
105B
8.8
8563
70.9%
4-5
707
5.9
9270
76.8%
5-6
523
4.3
9793
81.1%
6-7
413
3.4
10206
84.5%
7-8
315
2.6
10521
87.1%
8-9
268
2.2
10789
89.4%
9-10
171
1.4
10960
90.8%
>10
1113
9.2
12073
100.0%
Decumulative
Percent
89.6%
83.0%
81.0%
78.7%
75.8%
73.7%
71.5%
69.6%
67.7%
65.6%
64.1%
60.8%
57.7%
54.6%
51.3%
48.6%
37.8%
29.1%
23.2%
18.9%
15.5%
12.9%
10.6%
9.2%
0.0%
-------
Table B-20
Trip Distance (miles) Classes
00:15 Tuesday, August 11, 1992 1118
QBS LOWER UPPER
1
0.010
0.013
2
0.013
0.016
3
0.016
0.020
4
0.020
0.025
5
0.025
0.032
6
0.032
0.040
7
0.040
0.050
B
0.050
0.063
9
0.063
0.079
10
0.079
0.100
11
0.100
0.126
12
0.126
0.158
13
C.158
0.200
14
0.200
0.251
15
0.251
0.316
16
0.316
0.3S8
17
0.398
0.501
18
0.501
0.631
19
0.631
0.794
20
0.794
1.000
21
1.000
1.259
22
1.259
1.585
23
1.585
1.995
24
1.995
2.512
25
2.512
3.162
26
3.162
3.981
27
3.981
5.012
2B
5.012
6.310
29
6.310
7.943
30
7.943
10.000
31
10.000
12.589
32
12.589
15.849
33
15.849
19.953
34
19.953
25.119
35
25.119
31.623
36
31.623
39.811
37
39.811
50.119
38
50.119
53.096
39
63.096
79.433
40
79.433
100.000
PERCENT CUM DECUM
2.88435
2.88
97.12
0.32356
3.20
96.80
0,47148
3. 67
96.33
0 31432
3,98
96.02
0.52695
4.51
95.49
0.54544
5.06
94.94
0.55468
5.61
94.39
0.53619
6.15
93,65
0.60091
6.75
93.25
0.58242
7.33
92.67
0.81353
8.14
91.86
0.61940
8.76
91.24
0.81353
9.57
90.43
1.29426
10.66
89.14
1.83970
12.70
87.30
2.65323
15.35
84.65
2.41287
17.76
82.24
3.06924
20.83
79. 17
3.47601
24.31
75.69
4.11389
28.42
71.58
4.6C386
33.02
66.98
5.80568
33.83
61.17
6.80410
45.63
54.37
6.43432
52.06
47.94
7.6C839
59.67
40.33
7.65462
67.32
32.58
6.78562
74.11
25.89
6.12000
80.23
19.77
5.24175
85.47
14.53
4.20634
89.68
10.32
3.40205
93.08
6.92
2.66248
95.74
4.26
1.34048
97.08
2.92
1.33124
03.41
1.59
0.55468
98.96
1.04
0.29583
99.26
0.74
0.32356
99.58
0.42
0.13867
99.72
0.28
0.12018
99.84
0. 16
0.12018
99.96
0.04
COUNT
312
35
51
34
57
59
GO
58
65
63
88
67
88
140
199
287
261
332
376
44S
498
628
736
€96
823
828
734
662
567
455
368
288
145
144
60
32
35
15
13
13
-------
Figure B-21
PERCENT
IB H
00123456789 I 122334455*
----------05050505056
1 23456789 I - -- -- -- -- -0
01223344556
S050503050
Distance Driven per Day (miles)
-------
Table B-22
Distance Driven per Day (mites)
Cumulative
Cumulative
Decumu l at i ve
CTDUN
frequency
Percent
F requency
Percent
Percent
4-5
1
0.5
1
0.5%
99.5%
5-6
1
0.5
2
0.9%
99.1%
6-7
3
1.4
5
2.3%
97.7%
7-8
4
1.9
9
4.2%
95.8%
8-9
3
1.4
12
5.6%
94.4%
9-10
4
1.9
16
7.4%
92.6%
10-15
34
15.7
50
23.1%
76.9%
15-20
34
15.7
84
38.9%
61.1%
20-25
38
17.6
122
56.5%
43.5%
25-30
21
9.7
143
66.2%
33.8%
30-35
10
4.6
153
70.8%
29.2%
35-40
21
9.7
174
80.6%
19.4%
40-45
11
5.1
185
85.6%
14.4%
45-50
7
3.2
192
88.9%
11.1%
50-55
10
4.6
202
93.5%
6.5%
55-60
6
2.8
20B
96.3%
3.7%
>60
8
3.7
216
100.0%
0.0%
-------
Table B-23
Distance Driven per
Day (miles)
OBS LOWER UPPER
1
0.100
0.126
2
0.126
0.158
3
0.158
0.200
4
0.200
0.251
5
0.251
0.316
6
0.316
0.398
7
0.398
0.501
8
0.501
0.631
9
0.631
0.794
10
0.794
1.000
11
1.000
1.259
12
1.259
1.585
13
1.585
1.995
14
1.995
2.512
15
2.512
3.162
16
3.162
3.981
17
3.981
5.012
18
5.012
6.310
19
6.310
7.943
20
7.943
10.000
21
10.000
12.589
22
12.589
15.849
23
15.849
19.953
24
19.953
25.119
25
25.119
31.623
26
31.623
39.811
27
39.811
50.119
28
50.119
63.096
29
63.096
79.433
30
79.433
100.000
PERCENT CUM
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.0000 0.00
0.4630 0.46
0.4630 0.92
3.2407 4.16
3.2407 7.40
8.3333 15.73
10.6481 26.38
12.5000 38.88
18.5185 57.40
9.2593 66.66
13.4259 80.09
8.7963 88.89
7.4074 96.30
1.8519 98.15
1.8519 100.00
COUNT
1
1
7
7
18
23
27
40
20
29
19
16
4
4
16:18 Tuesday, August 11, 1992
DECUM
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
99.54
99.08
95.84
92.60
84.27
73.62
61.12
42.60
33.34
19.91
11.11
3.70
1.85
0.00
-------
Table B-24 Vehicle Operation (minutes per day)
16:18 Tuesday, August 11, 1992 11
OBS LOWER UPPER
1
0
0
2
0
10
3
10
20
4
20
30
5
30
40
6
40
50
7
50
60
8
60
70
9
70
80
10
80
90
11
90
100
12
100
110
13
110
120
14
120
130
15
130
140
16
140
150
17
150
160
18
160
170
19
170
180
20
180
190
21
190
200
22
200
210
23
210
220
24
220
230
25
230
240
26
240
250
27
250
260
28
260
270
29
270
280
30
280
290
31
290
300
PERCENT CUM DECUM
0.0000
0.00
100.00
0.0000
0.00
100.00
1.8519
1.85
98.15
6.9444
8.79
91.21
8.7963
17.59
82.41
10.1852
27.78
72.22
12.9630
40.74
59.26
18.5185
59.26
40.74
12.0370
71.30
28.70
8.3333
79.63
20.37
6.9444
86.57
13.43
2.7778
89.35
10.65
5.0926
94.44
5.56
1.3889
95.83
4.17
1.8519
97.68
2.32
0.4630
98.14
1.86
0.0000
98.14
1.86
0.4630
98.60
1.40
0.9259
99.53
0.47
0.0000
99.53
0.47
0.4630
99.99
0.01
0.0000
99.99
0.01
0.0000
99.99
0.01
0.0000
99.99
0.01
0.0000
99.99
0.01
0.0000
99.99
0.01
0.0000
99.99
0.01
0.0000
99.99
0.01
0.0000
99.99
0.01
0.0000
99.99
0.01
0.0000
99.99
0.01
COUNT
4
15
19
22
28
40
26
18
15
6
11
3
4
1
1
2
1
-------
Table B-25 Number of Trips per day
OBS LOWER UPPER
1
0
D
2
0
1
3
1
2
4
2
3
5
3
4
6
4
5
7
5
6
e
6
7
9
7
8
10
8
9
11
9
10
12
10
U
13
11
12
14
12
13
15
13
14
16
14
15
17
15
16
IB
16
17
19
17
18
20
18
19
21
19
20
PERCENT CUM DECUM
0.0000
0.00
100.00
0.0000
0.00
100.00
0.0000
0.00
100.00
4.1667
4.17
95.83
7.8704
12.04
87.96
16.2037
28.24
71.76
15.2778
43.52
56.48
11.1111
54.63
45.37
13.8889
68.52
31.48
12.5000
81.02
18.98
6.9444
87.96
12.04
3.7037
91.66
8.34
4.1667
95.83
4.17
0.9259
96.76
3.24
0.9259
97.69
2.31
0.4630
98.15
1.85
0.4630
98.61
1.39
0.4630
99.07
0.93
0.0000
99.07
0.93
0.4630
99.53
0.47
0.4630
99.99
0.01
CDUtiT
9
17
35
33
24
30
27
15
e
9
2
2
1
1
1
1
1
16:13 Tuesday, August 11, 1992 9
-------
Table B-26 Number
of Stops
per hour
OBS
LOWER
UPPER
COUNT
PERCENT
CUM
DECUM
1
0
0
0.0000
0.00
100.00
2
0
2
0.0000
0.00
100.00
3
2
4
0.0000
0.00
100.00
4
4
6
0.0000
0.00
100.00
5
6
8
0.0000
0.00
100.00
6
8
10
0.0000
0.00
100.00
7
10
12
0.0000
0.00
100.00
8
12
14
2
0.9259
0.93
99.07
9
14
16
1
0.4630
1.39
9B.61
10
16
!8
1
0.4630
1.85
96.15
11
18
20
3
1.3889
3.24
96.76
12
20
22
4
1.8519
5.09
94.91
13
22
24
6
2.7778
7.87
92.13
14
24
26
6
2.7778
10.65
89.35
15
25
28
10
4.6296
15.28
84.72
16
28
30
13
6.0185
21.30
78.70
17
30
32
8
3.7037
25.00
75.00
18
32
34
11
5.0926
30.09
69.91
19
34
36
19
8.7963
38.89
61.11
20
36
38
22
10.1852
49.08
50.92
21
38
40
23
10.6481
59.73
40.27
22
40
42
25
11.5741
71.30
28.70
23
42
44
12
5.5556
76.86
23.14
24
44
46
17
7.8704
84.73
15.27
25
46
48
10
4.6296
89.36
10.64
26
48
50
6
2.7778
92.14
7.86
27
50
52
6
2.7778
94.92
5.08
28
52
54
2
0.9259
95.85
4.15
29
54
56
4
1.8519
97.70
2.30
30
55
58
1
0.4630
98.16
1.84
31
58
60
4
1.8519
100.01
-0.01
16:18 Tuesday. August 11, 1992 10
-------
Table B-27 Speed Frequency Distribution
18:07 Monday, August 10, 1992 4
OBS LOWER UPPER
1
0
0
2
0
5
3
5
10
4
10
15
5
15
20
6
20
25
7
25
30
8
30
35
9
35
40
10
40
45
11
45
50
12
50
55
13
55
60
PERCENT CUM OECUM
19.0502
19.05
80.95
4.8026
23.85
76.15
4.8026
28.65
71.35
5.1761
33.83
66.17
8.9648
42.79
57.21
16.9691
59.76
40.24
15.9018
75.66
24.34
8.0043
83.66
16.34
4.4824
88.14
11.86
0.8538
88.99
11.01
2.8815
91.87
8.13
5.5496
97.42
2.58
2.5614
99.98
0.02
COUNT
357
90
90
97
168
318
298
150
84
16
54
104
48
-------
Tabic B-28 FTP Acceleration Frequency Distribution
DBS
LOWER
UPPER
COUNT
PERCENT
CUM
DECUM
1
-4
-3
104
5.5556
5.56
94 .44
2
-3
-2
79
4.2201
9.78
90.22
3
-2
-1
117
6.2500
16.03
83.97
4
-1
0
833
44.4979
60.53
39.47
5
0
1
430
22.9701
83.50
16.50
S
1
2
157
8.3868
91.89
8.11
7
2
3
71
3.7927
95.68
4.32
8
3
4
81
4.3269
100.01
-O.Ol
18:07 Monday, August 10, 1992
-------
Table B-29 FTP Power Frequency Distribution
18.07 Monday, August 10, 1992 7
08S LOWER UPPER
1
0
20
2
20
40
3
40
60
4
60
80
5
80
100
6
100
120
7
120
140
8
140
160
9
160
180
10
180
200
PERCENT
CUM
DECUM
31.5291
31.53
68.47
27.4696
59.00
41.00
18.1326
77.13
22.87
9.7429
86.87
13.13
6.4953
93.37
6.63
2.8417
96.21
3.79
1.3532
97.56
2.44
1.3532
98.91
1.09
0.2706
99.18
0.82
0.8119
99.99
0.01
COUNT
233
203
134
72
48
21
10
10
2
6
------- |