I^HIiMHllMHMH  ^^^^^ «Hli ^^Bi^BHiH^^^  HBH  1HH  ^^^^F      H^B W ^^B^r   ^^^fr    ^^^H  ^^^B^^H ^^B  ^^^^

-------
mOil^SllJl)*!?!]^ a logical, commonsense thought
process learned through experience that helps orga-
nize ideas, concepts, and information into a useful
form or solution. HOHIDIIIill^^rO©^ a logical,
commonsense thought process learned through
experience that  helps organize ideas,  concepts,
and information into  a useful form  or  solution.
mUMUlll)5!?!!   a logical, commonsense thought
process learned through experience that helps orga-
nize ideas, concepts, and information into a useful
         I     ml
form or solution. QQHi'iSJFIBOi^Il©* a logical,
commonsense thought process learned through
experience that helps organize ideas,  concepts,
and information into  a useful form or solution.
mO^S12i)^P2(02 a logical, commonsense thought
process learned through experience that helps orga-
nize ideas, concepts, and information into a useful

-------
HEURISTIC ROUTING
This publication (SW-113) was written
for the Federal solid waste management programs
by KENNETH A. SHUSTER and DENNIS A. SCHUR
U.S. ENVIRONMENTAL PROTECTION AGENCY
1974

-------
An environmental protection publication (SW-113) in the solid waste management series
                    For sale by the Superintendent of Documents, U.S. Government Printing Office
                                   Washington, D.C. 20402 - Price $1.05
                                      Stock Njmber 5502-00120

-------
  The collection of solid wastes generally represents from 70 to 80 per-
cent of a community's total waste management costs. To help reduce
these costs, and at the same time improve the quality of community
services, over the past several years EPA's Office of Solid Waste Man-
agement Programs has been conducting studies on solid waste storage
and collection systems. This report on heuristic routing is the first of a
series of reports documenting the studies. Others  will include:
      •  A Five-Stage Improvement Process for Solid
        Waste Collection Systems
      •  Management Information System for Residential Solid
        Waste Collection
      •  Policies and Methodologies for Solid  Waste Collection
      •  Districting and Route Balancing
  Additional projects are  under way or  planned that ultimately will
complement this information. These include studies on: rural storage
and collection systems; inner-city storage and collection systems; finan-
cial mechanisms for both capital and operating expenses; institutional
and organizational  arrangements, ranging from public collection sys-
tems to open  and  free private competition  working within various
regulatory and utility structures.
  All the studies and resultant reports have two major purposes:  (1) to
provide information or guidelines on the many possible alternatives for
storage  and collection systems; (2) to provide relatively simple, but
effective, solid  waste management tools for  evaluating  the systems.
These management tools are designed to measure the effectiveness and
efficiency  of the  various kinds of storage and collection systems, to
identify high cost centers in the systems, and to provide predictive infor-
mation on the effect of changes in the systems.
  The objectives of a storage and collection system should be to provide
service  that is  economical, that is environmentally sound, that is
aesthetically acceptable, and that assures a continuity in service. These
same objectives also are the underlying measures of storage and collec-
tion effectiveness used in  developing the guidelines and  the manage-
ment tools.

                                  -ARSEN J. DARNAY
                                    Deputy Assistant Administrator
                                    for Soh'd Waste Management
                                                                            111

-------

-------
                   INTRODUCTION  1
MICRO-ROUTING—WHY AND WHEN  5
             HEURISTIC APPROACH

              TO MICRO-ROUTING
 PROCEDURES FOR MICRO-ROUTING  13
    PATTERN METHOD OF ROUTING  16

       FACTORS TO CONSIDER FOR  19
                IMPLEMENTATION
  APPLYING HEURISTIC ROUTING TO
  HUNTINGTON WOODS. MICHIGAN
23
                     CONCLUSION 38

                       EXERCISES 38
                      REFERENCES 45
                       ABSTRACT 45
                          TABLES 24

                                  24

                                  30


                                  30
    what is the heuristic approach
    to problem solving?
    what is routing?
    why a heuristic approach to micro-routing?

    what is the value of micro-routing?
    when should it be used?
 6  what are the heuristic rules for micro-routing?

    what routing patterns should be applied?
    how are the heuristic rules applied?
    what preparations are needed for routing?
    what data are required for routing?
    routing, how is it done?
    evaluating existing policies
    compiling the data
    reviewing existing and alternative operations
    determining the new routes
    implementation considerations
    effectiveness of the micro-routing program
    table 1. seasonal variations in solid waste
    collection—City of Huntington Woods, Michigan

    table 2. residences serviced by day of week
    and truck [Huntington Woods]

    table 3. time required for various collection
    activities [Huntington Woods]

    table 4. old and new collection routes
    [Huntington Woods]

-------

-------
what  is  the

heuristic approach

to problem solving?

The heuristic approach to problem solving con-
sists of  applying  human  intelligence,  experi-
ence, common sense and certain rules of thumb
(or heuristics) to develop an acceptable, but
not necessarily an optimum, solution to a prob-
lem. Of course, determining what constitutes an
acceptable solution is part of the task of decid-
ing what approach to use;  but broadly defined,
an acceptable solution is one that is both reason-
ably good (close to optimum) and derived with-
in reasonable effort, time, and  cost constraints.
Often the  effort  (manpower, computer, and
other  resources) required, the time  limits on
when the solution is needed,  and the  cost  to
compile, process, and analyze all the data re-
quired for deterministic or other complicated
procedures  preclude their usefulness or favor
the faster, simpler heuristic approach.
  Thus, the heuristic approach generally is used
when deterministic techniques or mathematical
models are  not available, economical, or prac-
tical. Its use, according to Bowman and Fetter,
predominates in at least three circumstances:
1. Where problems are so complex that,  though
   the essence of the problem may be stated in
   a mathematical framework,  the computation
   required is  quite  unfeasible, even  on the
   largest computer.
2. Where  problems—especially policy prob-
   lems with which top managers must grapple
   —are so amorphous that a  mathematical
   model cannot capture their most important
   characteristics.
3. Where, although a mathematical model may
   be successfully employed, the prelude to the
   model and the work subsequent to the model-
   ing must be of a less formal nature. 1

  Most complex problems encountered in the
real world fall into one or more of the preceding
categories and often can be solved through the
use of heuristics. Or, frequently  the most eco-
nomical and practical approach  is a combina-
tion of rigid models and heuristics.
what  is routing?

 The term routing  has been applied to solid
 waste  management in several  different ways.
 As a result, several models or approaches to
 solid waste routing have been developed, each
 of which addresses  itself  to a very different
 problem. The problems,  or models, may be
 divided into  three  categories: macro-routing,
 districting and route balancing, and micro-rout-
 ing (Figure 1).
  Macro-routing determines the assignment of
 daily collection routes to  existing  processing
 and disposal sites. The objective is  to optimize
 the use of processing and disposal facilities in
 terms of the daily and long-range capacities and
 operating costs of the facilities, while minimiz-
 ing the round trip haul  time (and hence the
 hauling cost) from the collection  routes to the
 processing or disposal sites. Information essen-
 tial to  macro-routing includes haul  times from
 the routes to the various processing and dis-
 posal sites,  crew size and vehicle capacity,
 expected arrival, queue, and service times at
 the sites, and short- and long-range capacities
 and costs of the sites. Macro-routing may also
 be useful in determining which of several pro-
posed processing and disposal  sites or garage
locations is most economical, again by consider-
ing such factors as costs, site capacities, and
                                                                                     1

-------
                                                            DISTRICTING:
MACRO-ROUTING:

ASSIGNMENT OF COLLECTION
ROUTES TO PROCESSING &
DISPOSAL  SITES.
              To landfill
              To incinerator
                                                            DETERMINATION OF ROUTE
                                                            BOUNDARIES, ROUTE BALANCING.

                                                                   !•:•:•:•:•:•:•:•:<  Monday
                                                                           Tuesday
                                                                           Wednesday
                                                                           Thursday

                                                                           Truck #1

                                                      MICRO-ROUTING:
                                                      PATH OF COLLECTION
                                                      VEHICLE ON  THE ROUTE.
              Figure 1. Types of routing: (1) macro-routing is assigning collection routes to
              processing and disposal sites; (2) districting is determining route boundaries and
              route balancing; (3) micro-routing is determining the path of collection vehicles
              on the  route.  The heuristic  approach described in this  report  applies to
              micro-routing.

-------
round trip haul times.
  Districting and route balancing determines a
fair day's work and divides the collection areas
into  balanced routes so that  all crews have
equal workloads. This is  achieved through a
careful  evaluation of how the collection crew
spends its time.
  Micro-routing looks in detail at each daily
collection service area to determine the path
that  the  collection vehicle should follow as it
collects from each service on its route. The ob-
jective is to minimize the  driving time on the
collection route through minimizing the dead
distance (i.e., street  segments that have  no
services or that are traversed more than once),
backing  of vehicles,  U-turns, left turns, collec-
tion  on  major streets during rush hour  traffic
and other delay times.


why a heuristic

approach  to

micro-routing?

Traditionally  there  are three approaches  to
problem  solving:  deterministic, heuristic-deter-
ministic, and heuristic.
  The deterministic  approach would  seek  to
solve the micro-routing problem by developing
a mathematical model.  This model would  al-
ways determine the optimum route based upon
the required input data. To date, no completely
deterministic  models have  been developed.
That is, there are no models that consider all the
factors pertinent to routing and that guarantee
the optimum  solution through eliminating all
other possible  alternatives.  There  are three
major reasons for this. First, it is too difficult to
quantify all the pertinent factors. Second,  it
would be prohibitively costly and time-consum-
ing to investigate all the possible solutions, even
with a computer model. And third, the collec-
tion route itself is subject to so many dynamic
variables that the optimum solution may change
frequently.
  As  applied  here the  heuristic-deterministic
approach attempts to solve  the  micro-routing
problem by using a computer to examine many
possible alternatives  and select the best alter-
native based  on some heuristic algorithm.  In
applying any of the computer models currently
being marketed, the user  (typically a consul-
tant) must spend time becoming familiar with
the community, preparing  information for com-
puter input, and adjusting and debugging the
computer  program  to perform the  necessary
computations for each specific community. The
computer  input includes information required
to develop an elaborate node system defining
the street network of the community (Figure 2)
and the characteristics of each street segment;
e.g., the length of the street and the  number of
services between nodes; traffic patterns; and the
quantity of waste generated.
  The heuristic approach deals with  the micro-
routing problem by  using a manual  procedure
to develop acceptable collection routes without
examining many possible alternatives.  It is
certainly possible to develop a good  solution to
the micro-routing problem using either the heu-
ristic-deterministic computer models or the man-
ual heuristic approach.
  The manual heuristic approach, however, has
some distinct  advantages  over the  computer
models.  It is faster, less costly, more flexible,
and easier for local personnel to apply. The
manual heuristic  approach requires less data
preparation than the computer models by elim-
inating  several such time-consuming tasks as
preparing information for computer input, dev-
eloping a  node network,  and modifying the
computer program.
  The computer is a powerful and useful tool,
but  in the case of  micro-routing, a man can
often perform  the task more economically. For
example, it is easy for a person to read a map,
but difficult to program a  computer to do so.
Heuristic rules are relatively easy for an experi-
enced person  to  apply while considering a
variety of street and traffic peculiarities, com-
pared to the difficulty and cost of programming.
It is also relatively easy for a person to add or
modify  a  rule for a specific  situation,  or to
modify a route at a  later date, compared to re-
writing programs, punching new data cards,
and rerunning a  computer model.  Municipal
personnel  can  readily understand the basis for
routings done  manually—a real advantage in
implementing,  in  conducting daily operations,
and in making necessary modifications as con-
ditions change.
 537-034 O - 74 - 2

-------
•
o
m
in
z
o
o
m
en


O
oo
m
-o
3J
O
O
30
m
O
          Figure 2.  A computer model would require 162 nodes for Huntington Woods,

          Mich.

-------
what is the value of

micro-routing?

  The value of micro-routing becomes obvious
when one considers the repetitive nature of the
collection process: any time saved by reducing
collection route distance and travel delays be-
comes cumulative. Assume, for example, that a
community has 10 trucks, each with a three-man
crew that services 450 residences daily in about
4 hours of on-the-route collection time. If 24 min-
utes of collection time per crew can be saved
by  rerouting,  then  nine collection  routes can
service the same  areas without taking  more
than the original 4 hours by adding 50 services
to each route. The total savings represent one
collection vehicle and three man-years of labor
(one crew). To realize these savings, these 24
minutes  per route  must, of course, be  utilized
through  further rerouting and route expansion
or through a change in crew size, rather than
dismissing the collectors  24 minutes  earlier
each day. The larger the collection system (i.e.,
number  of crews) the greater is the absolute
savings potential.
when should it

be  used?
In many communities the  existing collection
routes have evolved as the communities have
grown, by sporadically adding a section here
and there to the various routes, or adding new
routes to the new areas. Rarely has the overall
route  structure  been periodically  examined.
Thus many communities have fragmented, over-
lapping, or unbalanced routes that have un-
necessary dead distance, delay times, and in-
equitable workloads. If this is the case, micro-
routing should be used.
  Rerouting is also needed whenever there is a
significant change in the collection system. Such
changes include: frequency of collection; point
of collection  (curbside, alley, or backyard);
crew size; truck size or equipment type; location
of disposal sites; type of storage container [209-
liter (55-gal)  drums,  cans, sacks, etc.]; com-
bined versus separate waste collection; or num-
ber of services. In case of a change in  disposal
sites, macro-routing should precede  micro-rout-
ing to determine round trip haul time for each
route.

-------
The heuristic approach to routing is a relatively
simple and expedient method for obtaining an
efficient  route layout that minimizes dead dis-
tances and delay times. The heuristic approach
could also be called the pattern method of rout-
ing since it relies heavily on the application of
specific  routing  patterns to certain  block con-
figurations. Admittedly,  efficient  routing re-
guires both skill and aptitude. But  guided by
certain heuristic rules and patterns, and through
experience, a router can readily develop the
ability to scan a map and rapidly and system-
atically plot timesaving routes.
what  are  the
heuristic  rules
for micro-routing?

  1. Routes should not be  fragmented or over-
lapping. Each route should be compact, con-
sisting of street segments clustered in the same
geographical area.
  2. Total collection plus haul times should be
reasonably  constant for each route in the com-
munity (egualized workloads).
  3. The collection  route should be started as
close  to the garage or motor pool as possible,
taking into  account  heavily traveled and one-
way streets. (See rules 4 and 5.)
  4. Heavily traveled streets should not be col-
lected during rush hours.
  5. In the case of one-way streets, it is beist to
start the route near  the upper end of the street,
working down it through  the looping process
(Figure 3).
  6. Services on dead end streets can be con-
sidered  as services  on the street  segment that
they intersect, since they can only be collected
by passing  down that street segment. To  keep
left turns at a minimum, collect the dead end
streets when they are to the right of the truck.
They must be collected by walking down, back-
ing down, or making a U-turn.
  7. When practical, steep hills should be col-
lected on both sides of the street while vehicle
is moving downhill for  safety, ease,  speed of
collection, and wear on vehicle, and to conserve
gas and oil.
  8. Higher elevations should be at the start of
the route.
  9. For collection from one side of the street
at a time, it is generally best to route with many
clockwise turns around blocks.
  Heuristic rules  8 and 9 emphasize the devel-
opment of a series of clockwise loops in order
to minimize left turns, which generally  are more
difficult  and time-consuming  than right turns
and, especially  for right-hand-drive  vehicles,
right turns are safer.
  10. For collection from both sides of the street
at the same time, it is generally best to route
with long, straight paths across the grid before
looping clockwise.
  11. For  certain block configurations within
the route, specific routing patterns should be
applied.
what routing
patterns
should  be  applied?
As the router gains experience, he will recognize
routings that are efficient for certain block pat-
terns.  Certain patterns  should be considered
whenever  the  grid has blocks  arranged as
shown in  the figures (Figures 3 through 7).
Seguential application of these patterns helps

-------
                             START
      	      4        '            •	'

Figure 3.  Specific  routing  pattern for one-way street, one-side-of-the-street
collection.  In  this pattern, collection  is made from both sides of the one-way
street during the pass. For wide or busy one-way streets, it is necessary to loop
back to the upper end and make a straight pass down the other side.

-------
STAR'
FINISH

>1
L


,.
,


1
(
,
1







J








u
I





J











J
f
1

\
\
I









^
1
— !— fc»










1




_

^




	


' 1

o
z
m
1
i



r 1

']
I,



;






f
i J

t





^
1
^





L ONEWAY 1




J
•— — — ^^





•^ 	



^

J



1 1




1
_













^

1

1



' (

— »—



"!




T1
_ i
j





1




i
I
,.f ,
, i
j





'
ONE SIDE OF
THE STREET
COLLECTION
f
START
FINISH

1
lj
I
1






















t






I

















—
1
O
m
1
1











	 ^»


•^ 	


	 ^«


-< 	



J

t




O








'

ME WAY









1
























i



-
<
'
BOTH SIDES OF
THE STREET
COLLECTION
r
    	^"   DIRECTION OF ONE-WAY STREETS

    Figure 4.  Specific routing patterns for multiple one-way  streets. Note  the
    one-way streets are paired with a clockwise movement.

-------
                                                         START •-
    START •

    FINISH
    FOUR-BLOCK CONFIGURATION APPLICABLE WHEREVER
    FOUR BLOCKS ARE POSITIONED AS SHOWN
                                                                   THREE-BLOCK CONFIGURATION
 START*
FINISH
                                                     START*
                                                    FINISH
    VARIATION OF THREE-BLOCK CONFIGURATION
VARIATION OF THREE-BLOCK CONFIGURATION
                Figure 5.  Some specific routing patterns for three- and four-block configura-
                tions. In all configurations, blocks may vary in size and shape.

-------
START a
FINISH '
i







i

i

>





j_



1
i

- <





t

i

r
P

0
at
ii-
a
                                                                    4x4 BLOCKS
                                                                    NO LEFT TURNS
                                                                    NO DEAD DISTANCE



J
START «
FINISH T

J



1 r
1 » ^

^
1 '
j i




\





\


*
i
, |


LJ.

•a-
'"'™ 	
nl^



i
» ^


1 '
i
j ,




\
t ,


                                                                    4x6 BLOCKS
                                                                    ONE LEFT TURN
                                                                    NO DEAD DISTANCE
                                                                    6x8 BLOCKS
                                                              ^.^ ONE LEFT TURN
                                                                    NO DEAD DISTANCE
                                START      FINISH
             Figure 6.  Combinations of the four-block pattern, one-side-of-the-street collec-
             tion. Note that each route is started midway on an evenly divided side of the
             grid, and uses the same routing pattern, with progression in a counter-clockwise
             fashion. For the larger grids, once the outside is routed the inside is routed in a
             clockwise progression.
10

-------
                      START
                                  FINISH
                                                     PAT TERN  A
                    FINISH
                                                     PATTERN   B

                 Figure 7.  In  specific  routing  patterns for both-sides-of-the-street  collection,
                 pattern A entails no left turns, and pattern B requires nine left turns. Dash lines
                 represent "dead distance" or non-collection segments of the route.
537-034 O - 74 - 3
                                                                                                            11

-------
yield efficient route layouts. After determining
which street segments have services and which
are without services, applying these  patterns
and developing additional patterns for on ef-
ficient route is relatively simple.
  To help simplify the problem,  a  grid simpli-
fication  technique, described in the  section
"Routing, How Is It Done?" may be applied be-
fore looking for block patterns.

how are the
heuristic rules
applied?

The heuristic  rules are  guides which, us&d in
conjunction with the patterns, help determine
the vehicle path or route. Because of the infinite
possibilities of street and block  arrangements
and routing patterns, a rigid procedure for rout-
ing cannot be applied and thus routing does not
lend itself well to being programmed for a
computer.
  In applying the heuristic rules and patterns,
the router should look ahead and behind sev-
eral  street segments and ask  himself these
questions:
•  Do any of the heuristic rules apply here?
  Are they being violated?
•  Have any street segments been left behind
  that will require a long dead distance to re-
  turn and pick up?
•  Are there already some long dead distances
  that might be  reduced through slight modi-
  fications?
•  Are there any peculiar  or unique character-
  istics of the area which should be considered?
•  Are there  any patterns  that can be  utilized
  in the routing?

  Once the initial layout has been determined,
it should be reviewed for further refinements.
The route should be  retraced and alternate
routings or modifications should be attempted
wherever long dead  distances are apparent.
Note,  however,  that to evaluate all  possible
routing alternatives is a formidable task that
even computer models do not attempt.
  In summary, the heuristic rules and  patterns
facilitate the routing process, but must be tem-
pered with a commonsense application and ex-
amination as the route is being developed.
12

-------
what preparations

are needed for

routing?

In  developing  cm efficient collection system,
micro-routing as  defined here is actually the
last of four tasks that might precede implemen-
tation. In order, these tasks are:
• Review and  evaluation of existing policies
  and methodologies
• Macro-routing
• Districting and route balancing
• Micro-routing
• Implementation
  A major tool that will assist in all of these
tasks  is a  management information  system
(MIS) based on  a daily collection activity re-
port.  The MIS  includes route, collection,  and
cost information records that help identify time
and cost centers. A report entitled Management
Information System for Residential Solid Waste
Collection provides further detail.2

  Collection policies and methodologies directly
affect  routing   and collection efficiency  and
should be reviewed, evaluated, and changed, if
necessary, before rerouting proceeds. This kind
of analysis is discussed fully in a report entitled
Policies and Methodologies for Solid Waste Col-
lection.3  Policies,  which primarily relate  to
levels of service, include:
•  Point of collection  (distance  from street to
  storage)
•  Frequency of  collection
•  Type and weight limits of storage device
•  Garden waste collection
•  Bulky waste (furniture, white goods) collec-
  tion
•  Separate versus combined collection
  Methodologies, or  on-the-route practices,  in-
clude :
• Crew size
• Type and capacity of collection vehicles
• Shuttle system
• Reservoir system
• One-side versus two-sides-of-the-street collec-
  tion
• Fixed lunch site and time
• Collection by drivers
• Incentive schemes
• U-turns and vehicle backing
• Filling vehicles before going to disposal site
  After policies and  methodologies have been
reviewed and revised and the collection areas
have been assigned to disposal facilities (macro-
routing), districting and route balancing must
be performed. The key to  districting and  route
balancing is determining  a  fair day's work—
the  reasonable number of services each day
for each collection area. Calculating a fair day's
work requires at least  a sound  estimate  of the
average time spent to  service a residence and
the amount of waste collected at the residence.
These figures can be used to derive the average
number of  residences that can be serviced  for
each service area, the number of  routes  re-
quired, and the size  of each  route to equalize
workloads. The route balancing process is de-
scribed more fully in a report entitled Districting
and Route Balancing for Solid Waste CoWection.4
  Once the equitable service numbers for each
route are  determined,  the community is first
divided into districts—one for each collection
day  in the week. Districts also should conform
to manmade and natural barriers such as major
streets, railroads, expressways, parks, rivers,
lakes,  gulleys,  and  mountains. Each district
then is divided into routes—one for each collec-
tion crew.
  Micro-routing  is performed after  districting
                                                                                       13

-------
and at the same time the balanced routes, are
developed. For example, once the service area
for Monday  is defined, the micro-routing pro-
cedure is applied in a clustered area until the
predetermined number of services is developed
into a route.  Then the  next route is developed
starting in the area which is still unrouted. This
procedure continues  until the whole  city is
routed. As the routing process continues, how-
ever, the  originally selected districts may have
to be modified to achieve more efficient routings.
what data
are required
for  routing?
All the information required for routing can be
recorded on community maps. First, indicate on
the community map(s) the number and type
(residential,  apartment,  commercial, institu-
tional, industrial) of services per street segment
for each side of the street. The remaining street
segments with no services on them are non-col-
lection  segments. Next, identify all one-way,
dead end, and heavily traveled streets. Indicate
which corner-lot residents  (if any) should be
asked to place their waste on a specific street
segment. Finally, indicate, for each street seg-
ment or service area, whether the crews are to
collect one or both sides of the street on a pass.
routing,
how is it done?
To help simplify the routing problem, it is often
helpful to reduce the grid by applying a grid
simplification technique  (Figure 8).  By  this
method, the  number of blocks in the grid are
reduced by combining blocks that have no serv-
ices on the common street segments which face
each other. The pattern and routing procedure
can then be applied to this simplified grid. Also,
by getting certain corner-lot residents to place
their wastes on specific streets, it may be pos-
sible to eliminate the need to traverse the ad-
joining  street.  By  encouraging  neighboring
residents to place  their wastes together, the
collection time can be speeded, and just as im-
portant, the stop-and-go wear  on the vehicle
will be reduced thus extending its life.
  Using the marked map, pick a starting point
and link the street segments in each district into
a continuous route by  applying the  heuristic
rules and patterns to systematically minimize
the dead distance and left turns. Terminate the
route when the number of services is about that
determined to  be reasonable for a fair day's
work.
  Determine Ihe starting point for the next route
and repeat the routing procedure, again apply-
ing the heuristic rules and patterns until all the
services within the district have been routed.
Continue until all districts have been routed. As
with districting, the service boundaries for each
route should be determined by considering nat-
ural and manmade boundaries.
  For areas with significant  seasonal fluctua-
tions in waste  generation, the most  common
solutions are to increase the crew sizes, extend
the work day (pay  overtime), or add a crew
to the reservoir area for peak periods. Another
alternative is to shorten each route (i.e., lessen
the collection task)  and add crews. This kind
of adjustment can be readily accomplished in
two ways:  (1)  by  designing one continuous
collection route for each day of the week or area
and then assigning crews to distinct sections of
the continuous route based on the previously
determined number of services per route; or (2)
by establishing different routes for each season
of the year. The second method is generally the
better of the two.
14

-------
     J
                Dept.of  Public Works
      o
      o
      o



1

X
4 1 Q
KINGSTON
10 10
LA SALLE


10 io|
BORGMAN |
1 0
TALBOTio

1 1
HART
1 0

8
NAOINEio
ELGIN;,
•1

6
5
3






_
X
1 2
1 0
1 0
9
1 0
10
10 K
10 f

10
1 0

8
10
p — 1
8 ;
5

'12 MS
LINCOLN c

( 8
1 LINCOLN 8\ \J
iVERNON7\V 2 	 •

9

\
OD *? P
I t SALLE LN'. 9
8
5 8
^ 5 8
I 9
s r
IALBUI LN. ' _
1 	 1 n
3 ULJ
1
1
1
1
1
\
^r^T'^ 98- g
: 	 .
• a
I 5 9

/ 10 7
6 2|| s**^^>- |
L -*^>: .
\ 	 ^\\ '^^^^n
1
1
(
r
Figure 8.  Grid simplification technique. Blocks with no services on common
street segments are combined  before routing. All of the daily grids in the
Huntington Woods example in this report are easily simplified by this technique.
This figure shows the Friday route.
                                                                                 15

-------
In the pattern method of routing, the first step
is to identify the garage and disposal site loca-
tions, heavily traveled streets, one-way streets,
and street  segments with services (Figures 9
and 10). The garage location is shown on each
example,  and  for simplification, assume  no
heavily traveled or  one-way streets exist and
all  street segments have services on them.
  The next step is to identify block patterns and
pick a starting point  nearest the garage. In the
four- and three-block patterns,  the solution is
obtained by connecting the patterns together as
well as possible through a trial-and-error (heu-
ristic)  process (Figure  9).
  In the 4X4 combination of the four-block con-
figuration, and  variations  of  the three-block
configuration, again  the solution is obtained
through a trial-and-error process of connecting
the patterns into a continuous route (Figure 10).
           GARAGE
                                                                          SAMPLE
                                                                          PROBLEM
                                                                          GRID
            Figure 9. Example  showing pattern method of routing. This solution has no
            dead distance and two left turns.
16

-------
GARAGE
                                                               PATTERNS
GARAGE
  FINISH


'














<


t


















J





'
















_-






1
J
)
1
I



<











                                                               PATTERNS
                                                               CONNECTED
                                                               INTO CONTINUOUS
                                                               ROUTE
      Figure 9.  Example showing pattern method of routing. This solution has no
      dead distance and two left turns.
                                                                                    17

-------
        GARAGE
       GARAGE
        GARAGE
          START V.^_*
          FINISH
                                   SAMPLE PROBLEM GRID
                                    PATTERNS IDENTIFIED
            Figure 10.  Example showing pattern method of routing. This solution has no
            dead distance and 10 left turns.
18

-------
   •M -|\   r\i* rl  pV B 1 (c£   '* N  f" 'i    f ^' f.V\\f v*^^ t • r 1" S'1 h cl
   «f 4-\ 11 -1  '«^Jd^l^)     i   * )    *• -*'x' Jdl'vl §)d Ji/dll-ji
  -4,*^ «i««8W,Mi^ '
-------
they are to place their waste on certain streets.
  All citizens in the community—even  those
whose collection day will not change—should
be informed that the solid waste collection serv-
ice is  being altered. Residents often become
accustomed to collection at a specific time of
day and may set their waste out accordingly, or
frequently they relate their own time of collec-
tion with that of their neighbors across  the
street or around the corner, which may have
changed. Thus informing everyone should mini-
mize missed collections.
  City officials also should  be apprised of the
proposed changes. Persuaded of these changes,
they often are important allies during the transi-
tion. Too, should they be approached by citi-
zens regarding  the  changes,  they should be
aware of what is taking place.
  The  most extensive education effort will be
directed toward the citizens. Several media are
available. One of the best is a letter from the
mayor or city manager  explaining the reasons
for the changes and how  the  changes riay
affect the citizens. Such  a letter uses a soft-sell
approach, implies endorsement by the city man-
agement for the new system, and may include
a telephone number to call for further informa-
tion (Figure  11).
  Another  good  way to notify the citizens is
through flyers (Figure 12). The letter and flyers
shown here are similar to those used in Kansas
City, Missouri, during their effective implemen-
tation program for a change that took place in
March 1971. Note that the flyer is simple, yet it
gives all pertinent information. The map helps
residents conceptualize the collection area  and
acts as a check to  assure  that the flyer  was
delivered in  the correct  area. In Kansas City,
the actual cards were color-coded by  day of
the week to assist in  answering phone inquiries
from the citizens.
  Other methods that can effectively inform the
public include  television and  radio  announce-
ments and articles and notices printed in local
newspapers.  Even with an initial  multifaceted
information  program, complete citizen under-
standing and cooperation requires time  and
patience. Interpersonal  contact may be bene-
ficial:  e.g., (1) extra office personnel and  per-
haps  extra  telephone  extensions  to  receive
complaints and questions during  the change-
over;  or  (2) personal interviews with specific
corner-lot residents.  Finally, using savings as
they accrue from the new collection operations
to upgrade the cleanliness and appearance of
the trucks, replace  unreliable trucks, provide
uniforms for the collectors, or  improve the dis-
posal  situation certainly helps sustain good
public acceptance.
  The next section is a step-by-step account of
how the heuristic routing procedure was  per-
formed in one of many communities that have
recently adopted this approach.
20

-------
                                                                                           Office of the Mayor

                                                                                           Julius C. Jones, Mayor
                            City of Anytown, USA
                            Heart of America
                                                               30th Floor, City Hall
                                                               Anytown, USA 64106
                            February 12, 1973
Councilmen at Large

Joseph Hall
1st District

James Smith
2nd District

William Glenn
3rd District

Thomas Rose
4th District

Mark Jones
5th District

Paul Hogan
6th District
District Councilmen

Robert Phillips
1 st District

Charles Connolly
2nd District

Alan Michaels
3rd District

Peter Robbins
4th District

William White
5th District

James Myers
6th District
Dear Citizen:

On Monday, March 5, 1973, we will introduce a new City-wide refuse collection system designed to
decrease the City refuse budget while still providing the same level of service to you, the citizens.

This savings is made possible through the thorough evaluation of our refuse collection system using
the latest of management tools and implementing improved collection techniques.

The collection service will continue to be once-a-week curbside collection. We have, however,
revised the City ordinance to permit the use of plastic sacks, which we encourage you to use. En-
closed is a brochure explaining the advantages to you and to the City through the use of plastic
sacks, and some helpful hints on bag usage and refuse storage.

The only change affecting the citizens directly is a change for some in their day of collection. We
have tried to minimize  the number of citizens that must change their day of collection.  This new
collection system means, however, that in most instances the time of day that the collectors collect
from each residence will change. The enclosed map shows the day of collection for each area. Please
note when your new day of collection will be starting March 5.

Please help our sanitation men provide service to you during this change over by having your waste
put out by 7:00 a.m., the time they start to collect.

If you have any questions or suggestions, please call 684-4311.

Thank you for your cooperation in this effort.

                                               Sincerely,
                                               Julius C. Jones
                                               Mayor

               Figure 11.  Sample letter from a mayor to citizens notifying them of a change in
               collection system.
                                                                                                                    21

-------
Important  Notice

Beginning March 5, 1973
Your New Refuse Collection Day will be

tuesday
 There will be no collection on
 fhe following holidays:
May 28, 1973
September 3, 1973
October 8, 1973
October 22, 1973
December 25, 1973
February 18, 1974
 If a holiday is observed on Monday or
Tuesday, your collection day for that
week will be WEDNESDAY.

For information call 254-7417
         39th
                 38th
            63rd St.
       Public Works Department
       Refuse Division
       Anytown, U.S.A.
                      Important  Notice

                      Beginning March 5, 1973
                      Your New Refuse Collection Day will be

                      Wednesday
                       There will be no collection on
                       the following holidays:
May 28, 1973
July 4, 1973
September 3, 7973
October 8, 1973
October 22, 1973
December 25, 1973
February 18, 1974
                       If a holiday is observed on Monday,
                       Tuesday, or Wednesday, your collection
                       day for that week will be THURSDAY.

                       For information call 254-7417
                                                       63rd St.
                            Public Works Department
                            Refuse Division
                            Anytown, U.S.A.
          Figure 12.  Flyers are effective media for informing citizens of changes in
          collection system.
 22

-------
The  Office of  Solid Waste Management Pro-
grams assisted the City of Huntington Woods,
Michigan, in applying a heuristic approach to
rerouting its solid waste collection vehicles. This
section briefly characterizes the community's
existing collection system at the time study was
initiated  and its  new collection system, and
traces the  development  of its new collection
routes.
  Huntington Woods is a  middle-class subur-
ban community just outside Detroit. The popula-
tion  (1970  census)  is 8,536, which represents
2,497 residences. From solid waste disposal rec-
ords the  average solid waste generation rate
was  estimated to be about 1.3 kg (2.9 Ib) per
capita per calendar day, or an average of about
31.5  kg (70 Ib) of solid waste each week from
each residence.  Seasonal  variations in the
amount of solid waste, which must be consid-
ered in designing an efficient collection route,
were also  determined before  the rerouting
study.


evaluating

existing  policies

A thorough review and evaluation of the city's
existing solid waste collection policies provides
the basis for a "before and after" comparison
that  reflects the increased efficiency of collec-
tion  system operations. Under both the former
and  the new systems, the city provides weekly
curbside  collection service to all residents and
to the city's few commercial services. About  80
percent of the  city's residents use plastic bags
for waste storage. The city's collection system
formerly  used two 12.2-cu meter  (16-cu-yd),
rear-loading packer trucks operated by two two-
man crews to collect one side of the street at a
time.
  The Huntington Woods city management per-
sonnel considered a rerouting program neces-
sary because they had acquired new collection
eguipment. The city had purchased a one-man,
side-loading collection vehicle to replace one of
the older rear-loading packer trucks. During a
trial period, the replacement was able to finish
the collection routes in  approximately 5 hours.
Thus, although the city had saved the equiva-
lent of 1 man-year in labor costs, the potential
for even greater savings through rerouting ap-
peared very promising.


compiling the data

The data required for rerouting the City of Hunt-
ington Woods included a series of three maps.
The first, prepared by the superintendent of the
Department of Public Works, was a detailed
street map of the city showing the number of
residential  services  for each street  segment
(Figure 13).  Note that  it  is important  to cor-
rectly identify  the number  of services on each
side of a particular  street  segment and the lo-
cation of the Department of  Public Works ga-
rage or motor pool. A  separate city map was
prepared to identify all one-way, heavily trav-
eled, and dead end streets  (Figure 14).
  A third map identified those street segments
where no waste would be placed for collection,
i.e.,  undeveloped segments  with  no services
and corner-lot  residences that do  not face the
street  (Figure  15).  By correctly  designating
these segments, they can  be eliminated from
the design of the collection  route and thus help
                                                                                      23

-------
minimize the dead distance in  the individual
routes. In Huntington Woods, many such street
segments were identified by  asking residents
of corner lots to place their  waste on a specific
street segment  (though in most cases  the cor-
ner-lot houses faced the desired  street segment
and the waste was already being placed there).
  Figure  16 shows the former collection routes
before rerouting and  a breakdown of service
areas by day  of  the week. A  comparison of
each truck's daily  workload shows definite im-
balances between the two trucks and in the
general workload from day to day, even though
service density and the amount of waste 'per
service are similar for each route  (Table 2).
On Monday, for example. Truck No. 1 serviced
281  residences  and Truck No. 2 serviced !!05
residences, a total of  586 services. On Friday,
however, these same  trucks collected 131  and
183 residences,  respectively, a total of only 314
services.  The old  collection routes  were  frag-
mented;  the Wednesday and  Thursday collec-
tion routes, for example, are typical  of  both
fragmented and unbalanced  routes  that  most
likely evolved with the development of the com-
munity (Figure 16 ).
reviewing existing

and alternative
operations
At this point in  the rerouting study the city
reviewed its existing collection system opera-
tions and identified possible alternatives. The
city management isolated four major areas of
the collection operation where possible alter-
natives existed. The areas, along with the alter-
natives, were:
1.  Equipment
   Side-loader, 22-cu-meter (29-cu-yd),
     (one available)
   Rear-loader, 12.2-cu-meter (16-cu-yd),
     (two available)
2.  Crew size
   One-man, side-loader only
   Two-man, driver and one collector
   Three-man, driver and two collectors
3.  Level of service
   Curbside, once a week
   Carryout, once a week (requires additional
     personnel)
                                      TABLE 1
                       SEASONAL VARIATIONS IN SOLID WASTE COLLECTION
                           CITY OF HUNTINGTON WOODS, MICHIGAN

                                         Solid waste to be collected
            Season
                                Average w t/day
                                                      Average wt/residence/wk

Winter
Spring
Summer
Fall
tonnes
10.6
15.0
18.2
19.2
tons
11.8
16.7
20.2
21.3
kg
21.2
30.1
36.4
38.3
Ib
47.2
66.8
80.8
85.2
                                       TABLE 2
                       RESIDENCES SERVICED BY DAY OF WEEK AND TRUCK
Residences serviced

Vehicle
Truck No. 1
Truck No. 2
TOTAL

Mon.
281
305
586

Tues.
289
267
556

Wed.
220
328
548

Thur.
323
170
493

Fri.
131
183
314
Daily
Average
248.8
250.6
499.4
24

-------
r
AVMHQIH 30011003
ir
Figure 13.  Detailed community map of Huntington Woods, Mich., showing
services per street segment.
                                                                              25

-------
                                        AVMHEIH 33011000





                      Figure 14.  Detailed community map showing special streets.
ir
26

-------
                        AVMHSIH 39011003





Figure 15.  Detailed community map showing non-collection street segments.
                                                                                   27

-------
                                  Figure 16.  Former collection routes.
28

-------
4.  Collection methodology
   Collection from one side of street
   Collection from both sides of street
   Plastic bags as mandatory storage
     containers
  The city  management  felt  that the existing
level of service (curbside, once  a week) and
collection policy (collecting from one side of
the street and voluntary  use of  plastic bags)
was the best of the alternatives considered. With
respect  to equipment and crew  size, the city
management wished to fully utilize the new one-
man collection vehicle and collect any remain-
ing waste using  a  rear-loading  packer truck
with a two-man crew. Thus the collection sys-
tem  was defined  by the  city, and all that  re-
mained was the design of the collection routes.
  Some additional data were required before
the actual  heuristic  routing  could be started.
The average round trip time to the disposal site,
average  time per service, average number of
services  per  street mile,  and average weight
per service are all important parameters in the
design of any efficient collection route and can
be obtained by using a daily collection activity
report. In Huntington Woods, the average round
trip time to the disposal site is 45 minutes; the
average time per service during peak generation
periods is 0.5 minutes and the average weight
per service during the  peak generation period
is 38.2 kg (85 Ib)  per residence per week.
  To determine the total  number of  collection
routes required, the total  number of residences
(2,500) was  divided by 5 (maximum number
of  possible collection days), which resulted in
a daily workload of 500  services. Thus, using
an average 0.5 minutes per service, an average
of  250 minutes per day, or about 4.25  hours per
day, would be required to collect waste from
500 residential services during the peak season.
The low collection time of only 0.5 minutes per
service is largely  due to the use of plastic bags
by 80 percent of  the residents. One-way bags
reduce collection time per service.
  Using the peak solid waste generation rate of
38.2 kg (85 Ib) per residential service, the total
weight collected from  500 services should  be
about 19.12 tonnes (21.25 tons)  per  day. The
city's one-man collection vehicle has a 22-cu-
meter (29-cu-yd)  capacity which can handle
payloads of  up to 7.2  tonnes (8 tons) or 326
kg/cu  meter  (550 Ib/cu yd). Thus during the
peak season, with an average of  19.2 tonnes
(21.25  tons) of solid waste to be collected and
a collection vehicle  with an average payload
of 7.2 tonnes (8 tons), three trips to the disposal
site are required. At the average 45 minutes per
trip, the total time required for three trips to the
disposal site is 2.25 hours per day.
  The  city policy allows two 15-minute coffee
breaks per day and a  30-minute lunch period.
The lunch break  is not paid time and can be
taken at the convenience of the collector. Times
are apportioned for the various  activities of the
collection operation,  based on  500 services per
day and the one-man collection vehicle (Table
3). The 1-hour variance allows for day-to-day
variations in waste generation, unusual delays
at the  disposal site,  and adverse weather con-
ditions. From the data analysis, the solid waste
collection  routes  for the  City of  Huntington
Woods should be designed to include approxi-
mately 500 residential services per route.
determining the

new routes

Because the number of services per collection
route is the single most important parameter in
designing an efficient collection route, at this
point in the routing study the extrapolated aver-
age  number of services should be verified. In
Huntington Woods,  the  existing  Wednesday
collection routes serviced  548 residences; there-
fore  the two were combined and selected as a
test route for the one-man collection vehicle. All
548 residences were adequately serviced in less
than the 8-hour work day.  Thus the proposed
500 services per route appeared to be a very
reasonable workload for the new collection sys-
tem configuration.
  The City of Huntington Woods was rerouted
manually using the heuristic approach (Figure
17).  A comparison of past and present collec-
tion routes dramatically depicts the changes in
service areas by day of the week and route
assignment (Figures 16 and 17). A comparison
of the performance of the  former collection sys-
tem  and the new rerouted  collection  system
shows that Truck No. 2, a rear-loading packer
                                                                                       29

-------
truck, has been completely eliminated and that
Truck No. 1, the one-man  collection vehicle, is
servicing the entire  city (Table 4). The new
collection routes are continuous, balanced, and
not fragmented or overlapping.
  New  street-by-street collection routes  \vere
developed for the Monday through Friday serv-
ice areas (Figures 18 through 22). The previ-
ously outlined heuristic rules and patterns, and
a systematic approach were the basis for the
design of all  the collection routes. Note that
since the route is continuous, the collector may
leave the route for a trip to  the disposal site
whenever his vehicle is fully loaded and simply
return  to the  point of  departure  to continue
collection.

implementation
considerations

Successful implementation of the proposed col-
lection routes depended  upon securing the sup-
port of the collection personnel and the citizens
           of the community.
             The collection personnel were  directly in-
           volved in and contributed to the rerouting study
           by providing much of the required data and by
           reviewing the proposed changes. The collector
           was given a complete route book  graphically
           describing the service area for each day of the
           week and street-by-street routing of his vehicles
           (Figures  18  through 22). The added respon-
           sibility of operating the new one-man collection
           vehicle  and the increased  number of services
           justified a wage raise. Displaced collection per-
           sonnel were reassigned to other areas of the
           Department  of Public Works.
            To balance the service areas into equal work-
           loads, only simple changes in  the service area
           boundaries were required, thus minimizing the
           areas of the city where the day of collection was
           changed because of rerouting (Figure 23).
           Only 275 residences, or  about 11 percent of the
           total residences in the city, were affected by a
           change  in the day of collection. A letter from
           the office  of the city manager informed these
           residents of the change in collection day and
           requested their cooperation.
                                       TABLE 3
                      TIME REQUIRED FOR VARIOUS COLLECTION ACTIVITIES
                      Activity
             Average time required
                   (hr)
                     Collection
                     Disposal
                     Coffee break (2)
                     Variance
                   4.25
                   2.25
                    .50
                   1.00
                        TOTAL
                                                         8.00
                                       TABLE 4
                    RESIDENCES SERVICED: OLD AND NEW COLLECTION ROUTES
               Route
            New collection routes:
               Truck No. 1
               Truck No. 2
                  TOTAL
                                                 Residences serviced
                                    Mon.
                                             Tues.
                                                      Wed.
501)
  I)
501)
492
  0
492
494
  0
494
                                                              Thur.
517
  0
517
                                                                       Fri.
Former collection routes:
Truck No. 1
Truck No. 2
TOTAL

281
3011
58li

289
267
556

220
328
548

323
170
493

131
183
314
494
  0
494
30

-------
    AVMHSIH 39011003





Figure 17.  New collection routes.

                                                                   31

-------

         C3
         a
         1  I
               LINCOLN8V\
             t VERNON 7V
\
TE
                                     START
                                     FINISH

                 Figure 18. Monday collection route: 500 residences; five left turns.
32

-------
       START
     x\\
Figure 19. Tuesday collection route: 492 residences; six left turns.
                                  33

-------
                    FINISH
                16
                T/ALBOT  15
                                1 5
 1 6
          16
,.
                   114
              /  HART    15
zsn
                                 1 7
                                      1 5
            114
       	Li
                                                             1 7
           F
      \Ji
                   ' START
             Figure 20.  Wednesday collection route: 494 residences; one left turn.
     ELEVEN MILE  ROAO
g
1
I;
l|
Q

1

14
KINGSTON 17
1 7
LA SALLE 15
BORGMAN 16
TALBOT 15
fl 4
\ HART 15

T
START





fcl
O


1 1
I 6
17
16
1 4
1 5
1 5
1 6
14
15







H



"£" "1
1 6 I
16 j
16
1 6
1 6 :
! j I
"• /i
a
"|
FINISH 1
)
4
1 1
14
1 1
1 7

5
— ^-^
6
1 1
^A
4Z
C3
h-
C9
Z
12^:
^
3C
1 4
1 3 1
1 0*
n
11
Jl
7/
                                                     1 rr» .-
             Figure 21. Thursday collection route: 517 residences; five left turns.
34

-------
Dept.of  Publ ic  Works
  FINISH
^  START
ELEVEN MILE  ROAD
                   IASALLE  IN.     9
Figure 22. Friday collection route: 494 residences; 16 left turns.
                                                             35

-------
                                        AVMHOIH 30011003
36
              Figure 23.  Area of Huntington Woods in which the  day of collection was
              changed because of rerouting.

-------
 effectiveness of  the
 micro-routing
 program

 Because micro-routing is the last system design
 step in the five-stage improvement process and
 other changes usually take place at the same
 time, it is difficult to isolate how  much of the
 resultant savings are  attributable solely  to
 micro-routing.
  In the case of Huntington Woods, for example,
 three separate changes ensued. First,  in 1970,
 two three-man crews were reduced to two two-
 man crews. Then, in August 1971, these crews
 were further reduced to one two-man crew and
 one one-man crew. Soon thereafter,  the heu-
 ristic routings were designed for the one-man
 crew with the two-man crew being eliminated.
 However, one of the two-man crew was retained
as a spare as explained below. The total work
 force therefore was reduced from six to two
men; the number of collection vehicles was re-
 duced from two to one.
  Since only the last of the three changes was
achieved predominantly through the  heuristic
routings, the effectiveness of the heuristic rout-
ing program is measured by the reduced costs
to the City of Huntington Woods from this third
change.
  From 1971 to 1972 the average hourly labor
wage increased from $4.42 to $4.77, or 7.9 per-
cent, and the total tonnage collected increased
from 3,886 torses (4,318 tons) to 4,062 tonnes
(4,513 tons), or 4.5 percent. But the total labor
charges decreased from $31,309 to $23,700, or
24.3 percent. These labor costs reflect the  elim-
ination of one man on the two-man crew, since
one individual  was retained to fill in for vaca-
tions, etc., and for street sweeping and a new
separate collection of newspapers. Since the
change occurred in the  fall of 1971 to distort
the 1971 figure, the actual savings in total labor
wage  reflects  a  28-percent reduction,  from
$32,738 to $23,700.  Over the 3-year period, of
course, the total savings to Huntington Woods
have been much higher.
  The effort  involved in rerouting  the City of
Huntington Woods was minimal in view of the
cost reductions. The public works superintendent
reported that he spent "one day off and on"
gathering the required information. The authors
of this report spent one day (two  man-days)
designing the routes  and preparing the route
book showing each service area and route. The
total effort was less than three man-days.
                                                                                    37

-------
To  see modern scientific  management tech-
niques applied to solid waste managemenl sys-
tems is encouraging. Yet, in the case of micro-
routing,  a less sophisticated, common-sense
approach—the heuristic approach—often is suf-
ficient. This approach has  the advantage  of
being a simple tool that can  be readily learned
and applied by local operating personnel.  In
addition, it avoids the high cost of existing com-
puter  models and  gives more  flexibility to the
routing—a necessity, since  communities  and
routes are so variable. The use of computers or
computer models is not to be discredited. Com-
puters have proved useful to solid waste man-
agement for billing systems,  for  storing and
manipulating information (the Management In-
formation System  for Solid  Waste Collection,
for example),  and  for models for more complex
procedures such as macro-routing.
  The major problem in routing today is that
very  few communities  or collection agencies
have ever studied  their routings and attempted
to improve them. Now that  a  straightforward,
heuristic approach has  been developed, more
communities can pursue the  lower  costs and
better service that potentially exist for them.
  Huntington  Woods was  a forerunner,  but
other communities have since used the heuristic
approach to routing their solid waste collection
vehicles:

  Akron, Ohio
  Beverly Hills, Mich.
  Birmingham, Mich.
  Clawson, Mich.
  Des Moines, Iowa
  East Peoria, 111.
  Fall River, Mass.
  Flint, Mich.
  Hot Springs, Ark.
  Huntsville, Ala.
  Lathrup Village, Mich.
  Little Rock, Ark.
  Portland, Maine
  Royal Oak, Mich.
  St. Petersburg, Fla.

  Each of these cities has realized substantial
savings or increased the level of service. Again,
not all of the savings necessarily can be at-
tributed to the heuristic routing alone.
Experience is the best teacher of routing pro-
cedures. To provide some practice for the reader,
this section contains three exercises to familiar-
ize the prospective router with the heuristic
routing techniques. See Figures 24, 25, and 26;
each figure is a map that includes all the neces-
sary  information and instructions to  develop
a route. Figure 24 shows a simple block con-
figuration for one route; Figure 25 shows the
same block configuration complicated by two
one-way streets.  Figure  26 shows a complex
street layout covering two routes. Figures  27
through 29 present solutions to the exercises.
38

-------
7
3 2
8

8
3 5
8

5


16 17


6

9
3 1
8

10
2 2
11









2


18 20

3

6
4 3
5









7
2 1
8

9
5 6
8

7
4 2
8










5
1 2
6

7
2 1
7

6
1 1
7









8
4 3
7

9
2 1
8

10
2 1
10
GARAGE




 Figure 24.  Exercise example: 373 residences; two-way streets.
                                                                         39

-------
7
3 2
8




8
3 5
8

5


16 17


6

g
3 1
8

10
2 2
11




f
>
2


18 20


3

6
4 3
5



7
2 1
8





V
>.
g
5 B
8

7
4 2
8
> in
UJ
Z
O







5
1 2
6

7
2 1
7

6
1
7
£.
O
I







8
4 3
7

g
2 1
8

10
2 1
10
                      GARAGE          *START




                      Figure 25.  Exercise example:  373 residences; one-way streets.
40

-------
Figure 26.  Exercise example:  997 residences; two collection routes.
                                                                               41

-------
             START
                        3                21

                                8
                        16
                                10
                                1 1
                        FINISH
                        I GARAGE
                                                 18         20
                                                                     2              '

                                                                           8
                                                                           1 0
              Figure 27.  Solution to exercise example,  Figure  24.  Includes 6 left turns, no
              dead distance.
42

-------
	 r
7
' 3 2
8
1 	 4 	 , — .

8
3 5
8

5



, 16 17


6
^
9
. 3 1
8
r ^
1,0
• 2 2
11
*


. t *

/


k



k j
1
1
>.

ci
f
'





> I

t
}
2



10 on
Jo /U

3
*
6
4 3
5


5
1 2
6
	 4 	
>»
7
2 1
7
— 	 k 	
6
1 1
7
	 «_, 	


1 r

\

,




'
X
o
>
0,
0
,1

j

r '



r '


	 r
7
2 1 '
8


9
5 6'
8
|r 	 < '
	 * 	 1
7
4 2 '
8

c
8
4 3 '
7
^ ;
9
2 1 •
8
/" 	 ^ 	
10
2 1 •
10
*. 	 ,4 ... — - —
     • GARAGE             s                   F

Figure 28.  Solution to exercise example, Figure 25. Includes 14 left turns; dash
lines represent dead distance.
                                                                                      43

-------
   ROUTE 1: 488 RESIDENCES, 17 LEFT TURNS.
   ROUTE 2: 509 RESIDENCES, 17 LEFT TURNS.
             CAN START AND FINISH AT
             EITHER POINT A OR B.
            Figure 29. Solution to  exercise example, Figure 26.  Route 1 serves 488
            residences; route 2 serves 509 residences. Each entails 17 left turns.
44

-------
              Shuster, Kenneth  A. and Dennis A. Schur. Heuristic Routing for SoJid
                 Waste  Collection Vehicles,  U.S. Environmental Protection Agency,
                 1974.
                 The development and successful application of a heuristic procedure
              for routing solid waste collection vehicles is described. Topics include
              (1) the value of routing, (2) reasons for rerouting,  (3) advantages of
              the heuristic approach, (4) heuristic rules, data requirements, and rout-
              ing  procedures, (5)  factors to consider when implementing this ap-
              proach. An actual example and exercises are included to show the rela-
              tive ease of application and to familiarize the reader with the heuristic
              approach to the routing of solid waste collection vehicles.
ya856
              1. Bowman, E. H., and R. B. Fetter. Analysis for production and opera-
                 tions management. 3d ed. Homewood, 111., Richard  D.  Irwin, Inc.,
                 1967. p. 452.
              2. Shuster, K. A. Management information system for residential solid
                 waste collection. (In preparation.)
              3. Shuster, K. A. Policies and methodologies for solid waste collection.
                 (In preparation.)
              4. Shuster, K. A., and D. A. Schur. Districting and route balancing for
                 solid waste collection. (In preparation.)
                  U. S. GOVERNMENT PRINTING OFFICE • 1974 O - 537-034
                                                                                         45

-------

-------