Data Management for Wide-area Responses: Technology Evaluation and Operational Expert Feedback Technical Summary Report by Timothy Boe1, Erin Silvestri1, Jamie Falik1 Matt Blaser2, Jim Mitchell2, Brian Cooper2, Leroy Mickelsen3, Lieutenant Commander Clifton Graham4, Katrina McConkey5, Molly Rodgers5 ^U.S. EPA Office of Research and Development (ORD) Center for Environmental Solutions and Emergency Response (CESER) Homeland Security and Materials Management Division (HSMMD) Durham, NC 27709 2U.S. EPA Region 5 3U.S. EPA Office of Land and Emergency Management (OLEM) Office of Emergency Management (OEM) Consequence Management Advisory Division (CMAD) 4U.S. Coast Guard (USCG) 5Eastern Research Group, Inc. (ERG) Morrisville, NC 27560 Contract EP-C-16-015 to Eastern Research Group, Inc. ------- Disclaimer The U.S. Environmental Protection Agency, through its Office of Research and Development, funded and managed the research described here under Contract EP-C-16-015 to Eastern Research Group, Inc. It has been subjected to the Agency's review and has been approved for publication. Note that approval does not signify that the contents necessarily reflect the views of the Agency. Mention of trade names, products, or services does not convey official EPA approval, endorsement, or recommendation. Contractor's role did not include establishing Agency policy. Questions concerning this document, or its application, should be addressed to: Timothy Boe U.S. Environmental Protection Agency Office of Research and Development Center for Environmental Solutions and Emergency Response 109 T.W. Alexander Dr. (MD-E-343-06) Research Triangle Park, NC 27711 Phone 919.541.2617 ------- Foreword The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the Nation's land, air, and water resources. Under a mandate of national environmental laws, the Agency strives to formulate and implement actions leading to a compatible balance between human activities and the ability of natural systems to support and nurture life. To meet this mandate, EPA's research program is providing data and technical support for solving environmental problems today and building a science knowledge base necessary to manage our ecological resources wisely, understand how pollutants affect our health, and prevent or reduce environmental risks in the future. The Center for Environmental Solutions and Emergency Response (CESER) within the Office of Research and Development (ORD) conducts applied, stakeholder-driven research and provides responsive technical support to help solve the Nation's environmental challenges. The Center's research focuses on innovative approaches to address environmental challenges associated with the built environment. We develop technologies and decision-support tools to help safeguard public water systems and groundwater, guide sustainable materials management, remediate sites from traditional contamination sources and emerging environmental stressors, and address potential threats from terrorism and natural disasters. CESER collaborates with both public and private sector partners to foster technologies that improve the effectiveness and reduce the cost of compliance, while anticipating emerging problems. We provide technical support to EPA regions and programs, states, tribal nations, and federal partners, and serve as the interagency liaison for EPA in homeland security research and technology. The Center is a leader in providing scientific solutions to protect human health and the environment. Through this effort, candidate tools were exercised and evaluated to assess the current state of technologies to enhance the U.S. Coast Guard (USCG) and EPA's ability to respond to and recover from a chemical, biological, radiological, and nuclear (CBRN) incident. The technologies and software recommended were exercised through a complete data management workflow during the Analysis for Coastal Operational Resiliency (AnCOR) field study that was held in May 2022. The operational considerations illuminated through this study provided invaluable information to ensure increased preparedness and, ultimately, more efficient, and successful field data acquisition and management activities. Gregory Sayles, Director Center for Environmental Solutions and Emergency Response ------- 1 OVERVIEW The U.S. Environmental Protection Agency (EPA) is designated as a coordinating agency, under the National Response Framework,1 with the authority to take actions to respond to releases of hazardous substances, pollutants, and contaminants, including leading the response. As such, EPA's role is to prepare for, respond to, and recover from threats to public health, welfare, or the environment posed by oil and hazardous materials incidents. Hazardous materials incidents can include accidental or intentional releases of chemical, biological, radiological, or nuclear (CBRN) substances. In the event of a CBRN incident, EPA can also have responsibilities to address debris and waste through decontamination, removal, and disposal operations. The U.S. Coast Guard (USCG) shares this responsibility for certain incidents in the maritime domain. The research conducted by this study aimed to streamline and improve the capabilities of USCG and EPA responders for a wide-area incident. Specifically, this project evaluated the current state of technologies for conducting site surveys (e.g., gathering information about the site or evidence that might inform sampling plan design) and managing sampling data following a wide-area incident, and correlated supportive technologies with specific field sampling activities to describe how tools and technologies are applied within an overall decision framework. Following a wide-area CBRN incident, from initial characterization sampling to evaluate the contamination event through clearance sampling and waste disposal processes, a substantial amount of data will need to be collected, checked for quality, and maintained to support decision-making. Depending on the size and scope of the hazardous contamination, data management could result in a significant technological undertaking that could continue for many years. Types of data that might be collected during the response include: Sample location, Sample matrix, Sampling method, Time and date of sample collection, Image of sample location or sampling surface, Sample collection personnel or team, Laboratory processing the analysis, Analysis results, Mapping data (e.g., Global Positioning System [GPS], light detection and ranging [LiDAR], photogrammetry), Documentation of quality assurance activities, and Decontamination method. 'The National Respoi nework is a guide to how the Nation responds to all types of disasters and emergencies. 1 ------- This project aimed to streamline capabilities and identify improved data management tools to better fit the needs of the Department of Homeland Security (DHS), USCG, and EPA responders following a wide-area contamination incident. This was accomplished by evaluating the current state of tools and technologies that facilitate the acquisition and subsequent management of field sampling data. This project had four (4) primary objectives: 1. Conduct a literature review and market research to identify relevant articles, reports, and other information describing research, ongoing initiatives by regional and state partners, and available commercial-off-the-shelf products that streamline and modernize field data collection activities; 2. Solicit subject-matter expert feedback from the response and research community on important functionality that field data acquisition and/or data management tools and technologies should have for responding to a wide-area incident; 3. Identify and evaluate technology to support response personnel based on recommendations provided by the response community; and 4. Conduct a field-scale demonstration to further evaluate operational aspects of selected technologies for the potential to enhance preparedness. This project was divided into two distinct phases: Phase 1 identified tools and technologies that were selected to be further evaluated and exercised during the planned demonstration day, are currently available commercial off-the-shelf (COTS) or government off-the-shelf (GOTS), and appear to have features that meet the largest number of needs [1], Tools recommended during Phase 1 for further evaluation included: Esri's ArcGIS Field Apps Suite - field data capture and mapping capabilities, RadResponder (CBRNResponder) - field data capture capabilities and aggregated reporting, Android Team Awareness Kit (ATAK) - increased situational awareness and offline communication needs, and EPA Scribe - storing field and laboratory data. Phase 2 of this project focused on exercising and evaluating the candidate tools recommended in Phase 1. Specifically, the project further evaluated the tools identified through this research during a technology demonstration day hosted by EPA's Homeland Security Research Program (HSRP), in association with the DHS/EPA-sponsored Analysis for Coastal Operational Resiliency (AnCOR) Data Project. AnCOR is a multi-agency program with the purpose of developing and demonstrating capabilities and strategic guidelines to prepare the U.S. for a wide-area release of a biological agent, including mitigating impacts to USCG facilities and assets [2], 2 ------- A secondary goal of this project was to document a repeatable, transparent, and stable workflow to support the AnCOR Wide Area Demonstration (WAD) data management needs. To address this need, the project team (consisting of federal responders, data management subject matter experts, and researchers) developed a Data Management Task/Workflow that identifies when and how various data management tools can be used across the response. Specific tasks that have a related data management component, the various tools that are available to support activities, and the established workflow among the tasks and tools were documented. The approach, observations, and recommendations resulting from the demonstration are summarized in the remainder of this technical summary report. A comprehensive review is available that describes the AnCOR Phase 2 project in its entirety [3], 2 APPROACH EPA conducted the AnCOR Data And Technology Assessment (DATA) demonstration (AnCOR DATA Demo) at EPA's Campus in Research Triangle Park (RTP), North Carolina on September 15th-l 7th, 2021. The AnCOR DATA Demo evaluated and operationally exercised tools and supportive technologies for use in sampling activities and data management workflow and processes following a contamination incident. The study took place outdoors (over a 200-acre test area) and involved approximately 20 volunteers from EPA and DHS/USCG. Participants completed a series of controlled tests using a variety of technologies to identify, locate, and document mock biological surface samples. As shown in Figure 1, 200 sampling points (point color indicates team association), five control points, and one test point were established to support exercising technology and software. Figure 1. Map of sampling, control, and test points throughout the RTP campus. 3 ------- Teams exercised different technology and field data collection software (consisting of three separate iterations) over two days. During each iteration, individuals were paired and assigned to teams to test different configurations of technology/software. Prior to beginning a sample collection phase, each team proceeded through a series of stations, as shown in Figure 2, and returned to report feedback following completion of the sampling activity. Figure 2. Progression of stations. Following the completion of an iteration, participants reported their observations and feedback. In addition to general observations and feedback, participants (including data managers) were asked to observe and report on their experiences using the technologies evaluated during the demonstration. For the Phase 2 evaluation, priority evaluation criteria were organized into the following four overarching categories: Software, Hardware, Technology/Software Configuration, and Operational Feasibility: Software Use of touch-sensitive data capture forms while wearing personal protective equipment (PPE); note: garden gloves were used as a substitute during testing, Toggling online or offline mode, Capturing an image using the camera feature, Capturing a video, Entering text using a finger and/or stylus, Using onscreen maps, Scanning a quick read (QR) code (for tracking samples), Synchronizing data from mobile application to centralized data storage, and Using navigational and geographic information system (GIS) features. Hardware GPS performance, Battery capacity, Capability to expand (USB/Bluetooth), and Ruggedness of device (i.e., performance related to excessive heat, light). Technology Software Configuration In advance of the exercise, AnCOR DATA Demo project team members were responsible for acquiring devices and configuring field data collection software. Observations related to the ease 4 ------- of implementing the acquired technologies and field data collection software were also captured and documented, including: Compatibility with other EPA systems, Database-compatible export formats, Ease of configurability, Ease of customization, Ease of maintenance, and Flexible data export formats. Operational Feasibility The AnCOR DATA Demo project team also evaluated several important feasibility considerations related to technologies and data management operations in the field. The following considerations were evaluated: Span of Control: The number of individuals or resources that one person can effectively manage during an incident according to the Incident Command System (ICS). Just-in-Time Training: Training personnel only when it is needed rather than in advance or on a predetermined frequency. These training opportunities are typically used as refresher courses prior to emergency response teams utilizing procedures or technologies. Offline Operation: Communications might be inoperable following a large-scale biological incident; therefore, the AnCOR DATA Demo evaluated the use of data acquisition tools without access to internet. PPE Limitations: The use of PPE can limit the dexterity of personnel, especially when using tablets that are touch-sensitive. Teams were randomly (via injects) asked to use thick garden gloves (to mimic PPE) and equipped with a stylus. Teams were then asked to interface with the software. Protecting Sensitive Equipment: The decontamination of expensive electronic equipment for reuse is essential. Previous field studies have successfully demonstrated that electronic tablets (e.g., Apple iPads) can be successfully decontaminated by encapsulating them in a water-resistant case and dunking them in a bleach solution [4], The use of external GPS equipment, however, complicates this process. Real-Time Quality Control Measures: Addressing secure, bulk data uploads (e.g., once an internet connection is re-established), storing and processing large quantities of data, and analyzing results in a collaborative platform are important operational considerations. In addition to evaluating the candidate software, EPA also maximized this opportunity to evaluate several other tools and technologies that can support implementing sample designs and data management needs during an event, including: MicroSAP - EPA's Sampling and Analysis Plan Template Tool is available to assist planners in developing a sampling and analysis plan (SAP) needed to collect data that are suitable for decision-making and/or determinations of existing conditions for all phases of a contamination incident involving pathogens in which EPA would be responsible for 5 ------- conducting sampling and analysis [5], While the project team did not use MicroSAP for the demonstration, the framework/guidance is a resource that can be used to support documentation of the sample collection and analysis procedures or methods to be used, sampling design, quality control procedures, and data reduction and visualization. Trade-Off Tool for Sampling (TOTS) - The Trade-Off Tool for Sampling (TOTS) allows users to create sampling designs and estimate the associated resource demand through interactive point-and-click tools to visually develop sampling plans for biological contamination incidents [6], TOTS was used by the AnCOR DATA Demo project team to create sampling designs for the AnCOR DATA Demo and to determine whether the proposed designs could be successfully executed within the allotted timeline of the demonstration. The sample designs were exported and used with field data capture software evaluated for the demo. Demo participants did not exercise TOTS. Simple QUIck REad Label (SQUIREL) - EPA's Simple QUIck REad Label (SQUIREL) tool is a lightweight tool that can be used to create sample label designs that are rendered in a QR code format that can be used with field data capture applications and/or laboratory systems; therefore, reducing human error when documenting sampling identifiers and other relevant information [7], Users can generate labels by entering study-specific nomenclature into text fields. SQUIREL was used by the AnCOR DATA Demo project team to create labels for sampling kits. Demo participants exercised capabilities related to scanning QR codes. 3 AnCOR DATA Demo Participant Observations and Findings A brief description of the software tools that were exercised during the demonstration and a summary of observations and feedback provided by the AnCOR DATA Demo participants is discussed in the sections below, organized by the following four primary categories: Software Evaluation, Hardware Evaluation, Technology/Software Configuration, and Operational Feasibility. 3.1 Software Evaluation As discussed, Phase 1 of this project identified software tools and technologies to further evaluate and exercise during the AnCOR DATA Demo. Emphasis was placed on available COTS and/or GOTS tools that appeared to have features that would meet the largest number of users' expressed needs. Field data acquisition software exercised during the AnCOR DATA Demo focused on: 1) ArcGIS Field Maps and 2) CBRNResponder2 2 While RadResponder was specifically identified during Phase 1, the Federal Emergency Management Agency (FEMA) sponsors a suite of "responder apps" collectively referred to as CBRNResponder that includes access to both RadResponder and ChemResponder (and soon, BioResponder). The demonstration evaluated CBRNResponder. 6 ------- The AnCOR DATA Demo project team had initially planned to evaluate ATAK as part of the AnCOR DATA Demo; however, the team experienced several impediments that prevented its full implementation. More information on ATAK can be found in the comprehensive project report [3], EPA's Scribe tool is routinely used by EPA's Environmental Response Team [8]; however, the AnCOR DATA Demo did not include a full data workflow that integrated analytical data communication to/from laboratories. Therefore, Scribe was not exercised during the demonstration. Data collected while exercising field data collection tools and software were instead stored on EPA's GeoPlatform for the AnCOR DATA Demo. 3.1.1 Esri's ArcGIS Field Apps Suite Esri provides a large suite of tools to support geospatially-driven tasks and analyses. EPA provides an enterprise-level offering for Esri's suite of tools. AnCOR DATA Demo participants only evaluated Esri's ArcGIS Field Maps tool during the exercise, whereas the project team utilized Esri routing tools for planning and the ArcGIS Dashboard and Tracking tools during the demonstration. 3.1.1.1 ArcGIS Pro Linear Referencing (Create Routes) The routing of teams for sampling or decontamination purposes is a significant challenge. Teams could encounter hazardous environments and/or spread contamination to otherwise clean areas. Because many of the tools evaluated in support of the AnCOR DATA Demo are GIS-based, the project used advanced geospatial capabilities [9] to determine optimal paths according to time and distance. Sample designs were transitioned directly from TOTS into ArcGIS Field Maps (TOTS automatically publishes sample designs to an ArcGIS web map with attributes to customize a data capture template). Using ArcGIS Pro's Linear Referencing (Create Routes) feature, samples were automatically grouped according to proximity, and paths were drawn specific to start and end locations. The results of this analysis were used to determine sample sequence (i.e., the sequence in which samples are collected). The resulting pathways could then be made available to other geospatial tools to support sampling team navigation. 3.1.1.2 ArcGIS Field Maps To support teams navigating to sampling locations, recording and documenting field data, and managing team assignments, a customized template for ArcGIS Field Maps was developed. ArcGIS Field Maps is an all-in-one application that uses data-driven maps to help mobile workers perform data collection and editing, find assets and information, and report their real- time locations. Field Maps supports both iOS and Android mobile operating systems and can operate in offline mode (for saving data locally) [10], Since Field Maps is based on a full-scale GIS platform, custom aerial imagery and feature layers (e.g., sampling points, paths, and boundaries) can be used to support situational awareness. Field Maps includes an integrated navigation capability that indicates the distance and bearing to the assigned sample. The tool visually prompts the user once they have reached the designated sampling location and documentation can begin. The tool further includes image and video 7 ------- capturing capabilities, a built-in QR code scanner, and conditional inputs for documenting sample types, sampling start/finish times, and observations. 3.1.1.3 ArcGIS Dashboards A customized ArcGIS dashboard was developed to show the status of sample collection activities in real-time. The dashboard is composed of three components: 1) Map to display the location and status (instantaneously updated when operating in online mode according to the surface type sampled); 2) Status counter to display the total number of samples plotted versus the total samples remaining to be sampled; and 3) Bag scans to display the sampling bag ID and sample location ID as they are scanned. This view can be used to detect erroneously scanned sample bags or locations. The dashboard can also be used to indicate instances where the sampling location diverged from its intended location to support real-time quality control corrections if appropriate. 3.1.1.4 ArcGIS Tracker In addition to optimizing navigation, the AnCOR DATA Demo also evaluated Esri's ArcGIS Tracker solution for tracking teams in real-time. Teams that were assigned cellular-activated tablets were tracked throughout the duration of their sampling activities. The functionality can be combined with the dashboard to improve situation awareness, health and safety, and the potential for plotting new samples in-situ depending on a given team's location. 3.1.1.5 ArcGIS Field Maps Data Storage Esri's ArcGIS Online (EPA's GeoPlatform) was used to store data collected using the Field Maps application. Data uploaded through Field Maps were automatically formatted into the appropriate schema and text entries were limited. This approach reduces the risk of generating inoperable data that break the schema specified by the data manager. Furthermore, ArcGIS Online supports adding imagery and video and associates those media files with sample data linked to specific geographic locations. 3.1.1.6 ArcGIS Field Maps Participant Observations The following are key findings for ArcGIS Field Maps based on user feedback and observations: Overall, Field Maps worked well, was easy to use, and was a useful data acquisition tool that participants are very likely to use again. Use of "edit" and "copy/copy all" in the software was confusing to participants. Logging information for each point was not intuitive. Users requested a more explicit/obvious user interface control to enable/start data collection for a sample point. Participants suggested incorporating a "done" button when selecting the collection time/date to denote completion of sampling (tracking sampling times was an ancillary data point and expected to be automated in the future). Sample finish time/date might not be necessary to collect. Participants suggested that it would be beneficial for a different color point or icon to display on the map when a sampling location status is updated to confirm data were successfully recorded. 8 ------- Older iPad devices appeared to more slowly record data and submission failed on several occasions. Entering notes on the form with both gloved and ungloved fingers was easy on all tablets evaluated. Issues regarding scanning QR codes were more likely associated with the label itself (printer/pixilation issue or misprints where the label was partially cut off) rather than an issue with the software feature and/or device camera. Participants suggested adding an option to enter sampler and/or team in the form (or using a QR code to denote the team collecting the sample). Field Maps was fully compatible with offline collection. Data captured were automatically uploaded to ArcGIS Online as soon as the tablets established an internet connection. All samples collected in offline mode were successfully captured and uploaded. 3.1.2 CBRNResponder While the AnCOR program was designed to address a biological agent contamination incident, Phase 1 research identified RadResponder (a collaborative tool for responding to radiological or nuclear emergencies) as a candidate tool to evaluate, along with a new tool, BioResponder, to assess potential applicability/expansion to support a biological event. The BioResponder tool was under development during the project study period but is expected to aid with the collection of biological samples and laboratory analyses [11], Following the completion of Phase 1 research, a single platform for accessing all CBRN event types was launchedCBRNResponder. CBRNResponder is a free application for emergency response organizations that is sponsored by the Federal Emergency Management Agency (FEMA) and other federal partners [12], At the time of the demonstration (September 2021), the full scope of data fields was unknown, and the applicability to the biological sampling events in this context (versus an epidemiological context) could not be assessed at that time. Nonetheless, the project evaluated the overall CBRNResponder framework that was available to determine potential applicability to support the AnCOR program. Similar to the Esri Field Apps suite, AnCOR DATA Demo participants only evaluated the CBRNResponder Field Data Collection application during the exercise, whereas the project team utilized the CBRNResponder website to access Event Management and Dashboards for planning and oversight during the demonstration. 3.1.2.1 Event Management and Configuration The AnCOR DATA Demo project team established an account with the CBRNResponder support team. A test event, DATA Day Test, was created to facilitate exercising the application. The project team evaluated the different data types that are included to support field data collection. For the demonstration, the team focused on the "Sample" and "Observation" data types. Prescribed data fields that are tied to specific data structures present some challenges given the need for flexibility to nimbly respond to changing data needs. Depending on the phase of an event, event type, and primary sampling objectives, what needs to be collected/characterized could change, and there does not appear to be a way to easily tailor 9 ------- sampling methods. Therefore, the AnCOR DATA Demo project team selected representative data types that most closely represented what data needed to be collected. Participants were instructed to ignore other irrelevant data fields and to only focus on a subset of fields to collect, including: Sample Type (Swipe), ID/Barcode (Scan sample bag QR Code), Surface Area/Units, Comment, and Photograph (via Attachment). A sample type of "swipe" (e.g., wipe sample) was selected for demonstration purposes. The types of sampling methods that will likely be needed for the AnCOR program (i.e., swab, microvac, aggressive air, or other innovative methods) do not currently align with the methods that are available for selection within the CBRNResponder application. The ability for users to add user-defined sample types or other "flexible" user-defined fields on an ad hoc basis is needed to meet EPA's needs. Once the field survey form to support data collection was defined, the project team defined the sampling locations for the event. A "Facility" was established to permit an association with sampling locations. The EPA RTP campus was used as a surrogate facility. Sampling locations can be entered one at a time (search/click on a map, enter latitude/longitude, or enter address) or pre-existing sample locations can be bulk uploaded via an Excel-based format with predefined coordinates using a provided template. While CBRNResponder allows shapefiles to be uploaded for reference during an event, there is not currently a feature that would allow a user to import/integrate geospatially-referenced sample locations from a map (e.g., using TOTS output with sample locations mapped). An additional data transformation would be required to generate and convert a shapefile to a comma-separated value (CSV) file to facilitate upload within CBRNResponder. For the demonstration exercise, going through this process (or manually transposing the coordinates) was not an issue. However, for a wide-area event that might require taking hundreds or thousands of sample points, the ability to rely on a single data source of geospatially- referenced sampling locations on a map is importantboth to minimize extra processing steps and to avoid data transformation errors. For the AnCOR DATA Demo, sample coordinates from TOTS output, along with sampling instructions, were manually entered into the provided template, and the template entries were bulk uploaded to the event. Once sample locations were established, the project team then created sampling team assignments to facilitate completing sampling and data collection tasks. Using the CBRNResponder website, assignments could be established one at a time or bulk uploaded by completing a provided template where both individual and team assignments could be made. For the AnCOR DATA Demo a single team was established, and assignments were associated with specific teams and automatically associated with predefined sample locations (latitude/longitude) and instructions. 10 ------- 3.1.2.2 Field Data Collection Several participants used the CBRNResponder application to assess its potential for wide-area incident response. Once data collection is complete, users submit data to an online data repository. Depending on whether the device is operating in on- or offline mode, data will synchronize once a connection is available. Data can then be viewed by event managers, as well as users through the application. Data can be exported from the platform; however, individual "data types" must be separately downloaded. 3.1.2.3 Event Dashboard and Map A preconfigured event dashboard is available to support event management. During the AnCOR DATA Demo, data managers could track field personnel using the map and predefined metrics could be arranged on an event dashboard to monitor a variety of metrics (e.g., data/type collected, field teams, assignments, organizational partners). Available features facilitate interacting with the map and tailoring views depending on the types of questions needing answers. For example, specific responders could be isolated to track status, sample status can be accessed, assignment information can be viewed, and additional GIS files can be viewed. Note, however, that GIS files must be uploaded to the platform and are not directly integrated with any ArcGIS online capabilities. Data captured within the platform can be exported into different formats, including keyhole markup language (KML), shapefile, and CSV. 3.1.2.4 CBRNResponder Data Storage CBRNResponder data are stored in a Microsoft Azure Cloud environment managed by vetted site administrators from the vendor organization that manages CBRNResponder for FEMA. Data uploaded are owned solely by the collecting organization and are only visible to other organizations if an organizational administrator provides explicit access [13], Uploaded data are formatted according to specific schema defined for various data types. Use of drop-down lists and limited free-form text entry fields is common among the different data types. Image files and documents can be uploaded and associated with a data point using the Attachments feature. Columns to display can be tailored and each data record can be expanded to view the full record, including any attachments associated with the data point. 3.1.2.5 Reports CBRNResponder offers a variety of preconfigured reports that can be generated [14], The AnCOR DATA Demo did not fully evaluate the range of offerings; however, two features that might be of interest include the ability to create barcodes and generating chain of custody forms. Barcodes can be generated for sample labels in advance of a sampling event and scanned using a device's camera. EPA also expressed the need for a feature to create chain of custody forms and electronic data deliverables to convey the information required for laboratory analyses. CBRNResponder can support creating "Sample Control Forms (SCF)" (chain of custody) and "Analytical Request Forms;" however, the ability to customize and/or tailor elements that are included on the form is currently unavailable. 11 ------- 3.1.2.6 CBRNResponder Participant Observations Two participants provided feedback on their experience using CBRNResponder during the AnCOR DATA Demo. The following are key findings based on user feedback and observations: Users noted that reliance on the device's default map (e.g., Apple Maps or Google Maps) was not as user-friendly as the integrated map offered by Esri's ArcGIS Field Maps. Toggling back and forth between the data collection application and a separate mapping application was frustrating for the users. Additionally, the default "mode of transport" required additional setting adjustments to ensure "walking" directions were enabled each time the map loaded. The mapping interface was cumbersome. Users expressed frustration at having to reload the map to view each sampling location where extra time was required to adjust settings and reorient the map. A participant suggested incorporating a yes/no button to confirm a status change rather than having to manually type "Yes" in "Confirm Status Change." Participants noted difficulty rendering QR codes; however, issues were likely a result of the label resolution. The application did not support videos, only images and documents. In the future, participants suggested correlating assignment descriptions with sample numbers. 3.2 Hardware Evaluation The AnCOR DATA Demo evaluated GPS and mobile devices to document any notable differences among devices that were used to support field sampling data collection activities. 3.2.1 Global Positioning System (GPS) Devices Sub-meter GPS receivers provide relative positional accuracy (hence the name, accuracy within 1 meter) and are used to determine the user's position, locate objects/areas of interest, and support navigation. Sub-meter GPS receivers may be used to aid in data acquisition and navigation when challenged with sampling designs that reference predefined sample locations. Several commercially-available sub-meter GPS receivers, including the Arrow Series GPS, SXblue, and Geode, were evaluated by a select group of participants in the AnCOR DATA Demo. The demonstration also evaluated the built-in GPS chip found in Apple's Wi-Fi + Cellular iPad models (e.g., iPad Air 2). A survey-grade receiver was used to capture control points in five separate locations. Once the control points and horizontal accuracy were determined, personnel equipped with the prescribed GPS devices (Arrow Series GPS, SXblue, Geode, iPad Air 2) navigated to each test location and captured five successive points (using a custom Field Maps template). The resulting points were then plotted in a GIS application to determine relative distance from the control point for each test location. Table 1 shows the average horizontal error for each device and location. 12 ------- Table 1. Average Horizontal Error (ft) by Location and Unit GPS Unit Adjacent to Forest Open Sky Near Tall Structure Field with Trees Dense Forest Average STD Geode 1.35 0.34 1.56 1.60 1.59 1.30 0.57 Arrow 1.07 0.91 10.27 5.43 2.05 3.95 4.07 iPad Air 2 11.78 5.29 8.26 4.14 8.41 7.58 2.74 SXblue 1.84 2.43 15.43 2.82 6.95 5.93 5.33 STD - standard deviation Overall, the Geode consistently outperformed the SXblue, Arrow, and iPad Air 2 (<3 ft of accuracy for this test condition) at all five locations. The Geode had the lowest average horizontal error (1.3 ft) and STD (0.57 ft). It should be noted that horizontal error can also be a condition of satellite geometry, signal blockage, and atmospheric conditions, which can vary by location and time of day/year. These conditions were not evaluated as part of this study. A control point (horizontal control) should be taken to assess conditions and accuracy prior to using a sub-meter GPS. During the AnCOR DATA Demo, built-in GPS on the tablets/phones or sub-meter GPS receivers (Arrow Series GPS, SXblue, Geode) were used to locate and navigate to predetermined sampling locations. The following key findings are based on user feedback and observations: Sub-meter GPS units worked better than cellular built-in GPS on all the devices evaluated, and the sub-meter units exhibited a better battery life. o Participants noted frustration with fluctuations of the location point on Samsung devices. Overall, with all receivers, but more so with the built-in GPS, signal seemed to be diffracted from local objects and would not stabilize or showed a decrease in accuracy when near solid structures or tree canopies. The Geode sub-meter GPS receiver mounted on a pole was reported as very accurate (<5 ft). The Arrow Series and SXblue sub-meter GPS receivers positioned on hats were more convenient than the Geode GPS on a pole. Sub-meter units need to be stabilized on the hat and face straight upward for best performance and accuracy. Avoid interference from multiple devices in close proximity where Bluetooth connections can sync with a nearby device. 3.2.2 Mobile Devices The AnCOR DATA Demo evaluated multiple electronic devices capable of documenting samples, uploading data, and connecting to sub-meter GPS units. The devices evaluated as part of this study included: Samsung - Galaxy Tab S7 tablet (Android, 11), Apple - 7.9-Inch iPad mini (5th Generation) with Wi-Fi (iOS, 10.3.4), 13 ------- Apple 10.2-Inch iPad Air 2 with Wi-Fi + Cell (iOS, 10.3.4), and Apple iPhone XR with Wi-Fi + Cell (iOS, 10.3.4). These electronic devices were evaluated based on general performance and user feedback. Generally, key issues centered on: Battery life, Screen brightness, Bluetooth connectivity, and Proper configuration (e.g., cellular data access, camera access, time out settings). The following are key findings based on user feedback and observations: Sampling events might require a longer battery life and use of an external battery, particularly when maintaining an active WiFi/cellular link to online data repositories, as well as reliance on GPS signals. Carrying an extra, fully-charged external battery would also impact sampling time requirements to avoid needing to return to the support zone (where personnel would need to also decontaminate and don/doff PPE to acquire additional supplies or charged equipment). Screen visibility was limited due to sun and accumulation of fingerprints (expected to be less of an issue when wearing protective gloves). Participants surmised that it would likely be difficult to see the screen with a full-face respirator (FFR). Additionally, increasing the screen brightness for better visibility rapidly drained the battery. Furnishing the tablets with a hand grip would make for easier operations in the field. Training on how to use the tablet's video feature is necessary. Overheating of devices could present issues such as lagging and decreased battery time. 3.3 Technology/Software Configuration Software considerations from the perspective of a data manager are summarized below. 3.3.1 ArcGIS Field Apps Suite EPA maintains an enterprise-level license for Esri products. Gaining access to the GeoPlatform (EPA's ArcGIS Online instance) for both the desktop-based application and mobile application was straightforward. Geoservices was responsive to inquiries and requests. Ample guidance and training materials are available on how to configure various tools for use. ArcGIS Field Maps directly integrates with EPA's TOTS tool output, as well as other geospatial assets that can be shared from EPA's GeoPlatform portal (e.g., surface classification/ characterization spatial analyses, building footprints, landcover, and operational zones). Field data capture forms can be easily tailored to meet stated data collection needs that are informed by site-specific data quality objectives. While not fully exercised during the AnCOR DATA Demo, EPA Regions have demonstrated creating sophisticated data entry validation measures to support 14 ------- real-time validation to prevent erroneous data entry and resolution of errors while still in the field [15, 16], Several other specific lessons-learned and best practices were noted: Use black text with a white outline when adding labels to features in maps to ensure visibility on different basemaps. Configure the original sampling map to the desired position for viewing sample points, enable the appropriate layers, and save the map. Doing so creates a default view for both the Dashboard map and Field Maps application map. Only create the "Offline View Area" in the offline map after all layers are added and properly configured. Overall, the project team found the suite of Esri ArcGIS Field applications easy to configure and tailor to meet the stated needs. Changes that could be easily incorporated as additional needs were identified. This provides important flexibility to EPA where conditions in the field might change, and the data management team needs to be able to quickly respond with any necessary adjustments. Using an online platform that can easily distribute and synchronize updates facilitates staying current with and responding to changing conditions. 3.3.2 CBRNResponder Gaining access to the platform (both the desktop-based application and mobile application) was straightforward, and the support team was responsive to inquiries and requests. Ample guidance and training materials are available to quickly begin using the tool. The underlying framework for the suite of applications is robust, including user administration (i.e., pre-establishing sample collection assignments and pushing notification of assignments to users' devices), team/event management, and a built-in operational dashboard. The administrative-related features provided are powerful, including push notifications, assignment status updates, and syncing capabilities. Tremendous benefit can be added by leveraging "routine management" functions without having to "build" the capacity for each event. The overall user administration and event tracking tools are very useful. This platform has potential; however, there are limitations related to the current alignment of relevant data fields and reporting needs. EPA's field data capture needs related to a biological contamination sampling event would best be met if the platform added data fields important to EPA biological sampling operations and/or provided users with the ability to define custom data fields and lookup values, as well as custom report options. Another beneficial enhancement would be better integration with ArcGIS Online and/or map services or layers that could be sourced from other online platforms. The need to separately upload (potentially large) geospatial data files, rather than incorporate by way of reference to an online URL, hinders EPA's ability to leverage other important operational geospatial assets. Additionally, unlike ArcGIS Field Maps, the application cannot directly integrate an operationalized map generated through EPA's TOTS tool. For a wide-area event that could require hundreds or thousands of sample points, the ability to rely on an authoritative/single data 15 ------- source of geospatially-referenced sampling locations on a map is importantboth to minimize extra processing steps and to avoid data transformation errors. Furthermore, the ability to navigate to geospatially-referenced sampling locations in real-time (e.g., heading, distance) is essential to implementing probabilistic sampling designs. CBRNResponder's navigation capabilities are limited to vehicle navigation using standard mobile routing platforms (Google/Apple Maps). Overall, the platform is easy to use and offers many robust features for the use cases on which design decisions were made. Enhancements related to custom data types, reporting, and better integration with other commercial GIS platforms would expand CBRNResponder's usefulness to meet EPA's AnCOR program needs. 3.4 Operational Feasibility Considerations Several important operational feasibility considerations were reinforced through the AnCOR DATA Demo. Observations and recommendations are listed below: Span of Control: For data mangers actively managing personnel who are equipped with newly introduced hardware and/or software, the span of control is significantly reduced. Troubleshooting might be required to assist inexperienced personnel and/or resolve hardware issues. It is recommended that data managers be associated with no more than four teams during an active response that requires the use of data acquisition tools and GPS hardware. The span of control could be increased through routine training of participants. Just-in-time Training: Just-in-time training was provided to AnCOR DATA Demo participants prior to entering the field. Participants who had never interacted with the prescribed technologies received the same level of training as those who were considered experts (through routine use). Following the demo, feedback and observations clearly determined that just-in-time training was inadequate for personnel who had never interacted with the tools and technologies that were exercised. Routine training should be provided to both emergency response and surge capacity personnel (including researchers) on a regular basis. Offline Operation: Teams were randomly chosen to operate in full offline mode (both field data collection applications exercised have offline data collection modes). Once the tablets resumed internet connectivity, results were automatically uploaded to the dashboard. All the samples collected in offline mode were successfully captured and uploaded. PPE Limitations: Overall, teams found devices and data capture forms easy to navigate using a stylus while wearing limited PPE (garden gloves to simulate nitrile gloves). Protecting Sensitive Equipment: While previous field studies have demonstrated that electronic tablets (e.g., iPads) can be successfully decontaminated by encapsulating them in a water-resistant case and dunking them in a bleach solution, there are currently no water-resistant cases available for GPS systems. Furthermore, the GPS systems typically consist of multiple parts (e.g., control unit, wires, and antenna). The current recommendation would be to keep the GPS system in the field (i.e., hot zone) for the 16 ------- duration of the daily sampling activity (<12 hours). The GPS antenna and control unit can be contained in Ziplock bags. The bags should be sealed using a durable adhesive tape. The wire portion can remain exposed. At the conclusion of the sampling day or every 12 hours, the GPS equipment should be removed, decontaminated, and charged (or charged in the field if conditions allow). The GPS equipment should be routinely checked against known control points to ensure operability. The steps to protect GPS equipment were not evaluated as part of the AnCOR DATA Demo but should be evaluated as part of the AnCOR wide-area field demonstration. Real-Time Quality Control Measures: Examples of real-time corrections that were evaluated include: 1) Reporting an incorrect sample method based on sample bag QR code (e.g., swab versus microvac), 2) Reporting an incorrect sample matrix, and 3) Capturing the sample at a distance outside of an established threshold designated by the sample design. Data managers and quality assurance (QA) participants exercised field data capture software capable of capturing data while in the field and submitting and synchronizing data to a consolidated online platform. The AnCOR DATA Demo evaluated data using Esri's suite with a two-step process: 1) Error Checking: the Field Maps form was designed to prevent users from entering erroneous data by confirming that input fields contained expected characters or limited free-text entries. If an error in the input menu was discovered or left blank, the tool would notify the user; and 2) Remote Review: since the AnCOR DATA Demo featured the real-time collection of data, the dashboard and associated data layers presented a unique opportunity to review data while sampling was underway. During the demo, two remotely located EPA personnel were asked to review data as they were uploaded to the cloud. Reviewers could either monitor the dashboard or the raw data. While data captured using CBRNResponder were not explicitly reviewed by a dedicated team during the collection event, the applications dashboard and data collection views would enable real-time data review. CBRNResponder also makes use of controlled data entry elements; however, there appears to be less flexibility to customize data validation rules as users are bound by what the application offers out-of-the box. A feature not exercised, but advertised as forthcoming, is a Chat function. The feature would presumably be very useful to quickly communicate/chat with a responder to convey any issues that might need correction/resolution while out in the field. 4 CONCLUSIONS AND RECOMMENDATIONS Through this project, EPA gained invaluable experience in understanding how to apply advances in technologies and software to improve field data acquisition tasks. Important technological issues were identified to inform future planning and training efforts. Based on the expressed needs of EPA and DHS/USCG and the experiences of participants in the AnCOR DATA Demo, the project team recommended using Esri's suite of tools and ArcGIS Field Maps to support field data acquisition efforts for the AnCOR program (and potentially future biological contamination 17 ------- sampling events). Consistent with the findings from a related effort to assess data visualization and analysis tools, the Esri suite has the most features that meet the largest number of needs, is familiar to and accepted by target stakeholders, and is generally viewed as easy to customize and tailor to meet the specific needs of the operation. Additionally, the Esri product suite is widely adopted among the response community and has been used by the USCG in support of various missions including search and rescue, pollution response, and response to natural disasters [17], CBRNResponder, and a forthcoming BioResponder, offer many promising features. At present, however, several key requirements for EPA's AnCOR program cannot currently be met namely, alignment with required data fields/types that will be collected and integration with real- time geospatial assets. The project team recommends that EPA continue engagement attempts with FEMA to convey EPA's needs regarding biological sampling (and other agents), and closely monitor FEMA's progress and tool enhancements (i.e., CBRNResponder and BioResponder) to determine whether these tools could better meet EPA's needs in the future. In pursuit of an additional project goal to document a repeatable, transparent, and stable workflow to support the AnCOR Wide Area Demonstration (WAD) data management needs, the project team also developed a Data Management Task/Workflow that identifies when and how various data management tools can be used across the response. Figure 3 illustrates specific tasks that have a related data management component that require input and support from the data management team. Tools available to support activities and the established workflow among the tasks and tools are illustrated, and specific tools recommended to support the AnCOR WAD are highlighted in blue. 18 ------- Design Sampling Plan Guidance Biological Sampling Framework Data Quality Objectives Methods Visual Sampling Plan (Probabilistic) Judgmental (Targeted) Tools MicroSAP TOTS Configure TOTS Export Geospatially-referenced sample locations Configure Access Privileges Figure 3. AnCOR Data management tasks and supporting tools. Recommendations resulting from Phase 1 of this study emphasized the need to create "a well- documented workflow, articulating desirable decision-making driven features, and defining required metadata and features needed to support data workflows." Further, enabling the response community to quickly adapt to new technology implemented using proven workflows will advance preparedness levels [1], Additional considerations resulting from experiences gained through completing Phase 2 of the project generally centered on the following topics: Quality Control Procedures/Objectives, Field Data Capture Form, Training, Operational Logistics, and Managing Devices. 19 ------- Table 2 summarizes important observations and feedback and identifies several actionable outcomes to further advance preparedness. Based on input from the project team, actions that should be prioritized are designated accordingly. Table 2. Additional Observations and Feedback Topic Observations/Feedback Quality Control Procedures /Objectives Closely integrate sampling and data management plans to ensure the right data are collected to inform decisions. Relate/define appropriate validation and quality control checks that would be required based on identified data needs. Compare sample matrix to sample method (via QR code scan); prevent data entry unless appropriate match. Check acceptable proximity to pre-established sampling point. Automate as many checks as possible through smart forms to prevent the entry of erroneous data from the start. Identify real-time checks (via a checklist) for data management team monitors to assess during the event and define the process for communicating any issues that might arise. Consider creating specialized views/queries that would support quickly identifying questionable entries. Field Data Capture Form Optimize forms to facilitate completion in less than one minute. Minimize free-form text entries. Maximize the use of "auto-collected" data to require fewer entries by a user (e.g., individual/team performing entry, day/time of entry, location of entry, sample status following entry). Incorporate visual cues on the digital map to illustrate the status of sample points. Training Clearly define specific objectives related to sampling event and data collection to ensure participants understand what, why, and how for each task they are asked to perform. Provide routine training to both emergency response personnel and surge capacity personnel (including researchers) for both data capture software and technologies that would be used during a response. Offer interactive, step-by-step training (either in person dry-runs or via PowerPoint or MS Teams). Train participants in advance on both the equipment (technology) and the software that they will use while in the field. Provide a classroom-based exercise to demonstrate tasks, discuss common ""gotchas." and answer questions well in advance of the exercise. Strategically pair teams to ensure whoever is charged with collecting data has the appropriate skills and training to successfully accomplish the tasks. Distribute laminated instruction cards and electronic versions that are pre-loaded on devices that provide quick tips and troubleshooting solutions to sample teams. Identify potential issues and corresponding courses of action (e.g., if device overheats, if battery capacity dips below 10%). Operational Logistics Optimize sampling routes to ensure that distances in-between samples are reasonable and make sure sampling routes avoid non-access areas or work zones. Plan sampling routes to avoid spreading contamination to otherwise clean areas. 20 ------- Table 2. Additional Observations and Feedback Topic Observations/Feedback Prepare QR code labels (for samples or team personnel identification) using high quality printers and labels to support optimal recognition by barcode readers/device cameras. Regarding extending battery life, determine whether samplers can operate in an offline mode for data submission where checkpoints are established (e.g., after completing five samples) to synchronize data or if they should carry battery backups/chargers. Associate data managers with no more than four teams during an active response that requires the use of data acquisition tools and GPS hardware. Leverage real-time location tracking among teams where teams near one another can provide troubleshooting support or assist with resolving immediately known collection errors/conflicts/issues from another team (as long as contamination spreading is avoided). Develop and execute a checklist for configuring and testing all hardware and devices that will be used in advance of the exercise. Implement all operating system upgrades, application updates, device settings (e.g., cellular, WiFi, Bluetooth, application access to camera, location services) to ensure optimal performance and configurations to support the associated data capture form features. Provide extra battery packs (adequately protected from contamination) to extend the capacity of a device's onboard batteries. Managing Devices Consider distributing WiFi source/extenders in the field to support device connectivity. Place and retain a GPS system in the field (i.e., hot zone) for the duration of the daily sampling activity (<12 hours). Consider distributing alcohol wipes to clean screens when in the field. Provide device hand grips to improve usability in the field. Consider having replacement devices on-hand and ready to activate should devices in the field begin to fail (e.g., battery needs recharging, device needs to cool down). Issue and attach a stylus for data entry. Design and implement measures to protect all sensitive equipment that might require decontamination. Several additional issues that were identified that require more information and/or research include: Address and document internal QA procedures that will occur prior to transferring samples for analysis by the laboratory; Define data auditing procedures and document explicit rules that will govern QA activities; Determine how chain of custody forms will be generated (i.e., what tool will be used) to generate required documentation; Address how chain of custody forms will be shared with labs; and Evaluate the impacts of weather (e.g., cold/hot temperatures, precipitation) on the usability and performance of tablets and GPS units. 21 ------- Through this effort, candidate tools were exercised and evaluated to assess the current state of technologies to enhance the USCG and EPA's ability to respond to and recover from a CBRN incident. The technologies and software recommended were exercised through a complete data management workflow during the AnCOR WAD held in May 2022. The operational considerations illuminated through this study provided invaluable information to ensure increased preparedness and, ultimately, more efficient and successful field data acquisition and management activities. 22 ------- 5 REFERENCES 1. U.S. EPA. 2021. "Data Management for Wide-Area Responses: Literature Review and Operational Expert Feedback." EPA/600/R-21/095. 2. U.S. EPA, ORD. 2019. "Analysis for Coastal Operational Resiliency." Reports and Assessments, https://www.epa.gov/emergency-response-research/analysis-coastal- operational-resiliencv. 3. U.S. EPA. 2022. "Data Management for Wide-area Responses: Technology Evaluation and Operational Expert Feedback." EPA/600/R-22/065. 4. U.S. EPA. 2013. "Bio-Response Operational Testing and Evaluation (BOTE) Project - Phase 1: Decontamination Assessment." EPA/600/R-13/168. https://cfpub.epa.gov/si/si_public_record_report.cfm?Lab=NHSRC&dirEntryId=263911 . 5. Silvestri, E., J. Cuddeback, K. Hall, T. Haxton, C. Jones, and J. Falik. 2021. "Sampling and Analysis Plan (SAP) Template Tool for Addressing Environmental Contamination by Pathogens." EPA/600/R-21/144. U.S. EPA, Washington, D.C. https://www.epa.eov/esam/sampline-and-analysis-plan-sap4emplate4ool-addressine- environmental-contamination-pathogens. 6. Trade-off Tool for Sampling (TOTS), n.d. U.S. EPA, Washington, D.C. Accessed March 14, 2022. https://tots.epa.eov/. 7. EPA QR Tool (version 1.3). n.d. U.S. EPA, Washington, D.C. Accessed March 14, 2022. https:// github. com/U SEP A/QR T ool. 8. U.S. EPA, OLEM. 2015. "SCRIBE Environmental Data Management." Overviews and Factsheets. https://www.epa.eov/ert/environmental-response-team-information-manaeement. 9. "Create Routes (Linear Referencing)ArcGIS Pro | Documentation." n.d. Accessed March 14, 2022. https://pro.arceis.com/en/pro-app/latest/tool-reference/linear-referencine/create- routes.htm. 10. "What's NewArcGIS Field Maps | Documentation." n.d. Accessed March 14, 2022. https://doc.arcgis.com/en/field-maps/faq/whats-new.htm. 11. "About | BioResponder." n.d. Accessed March 14, 2022. https://www.bioresponder.net/#about/index. 12. "About | CBRNResponder." n.d. Accessed March 14, 2022. https://www.cbrnresponder.net/#about/index. 23 ------- 13. "CBRNResponder | Terms of Use." n.d. Accessed March 14, 2022. https://www.cbrnresponder.net/#account/request. 14. "CBRNResponder Bonus Webinar: Event Reports." 2020. https://YOiitu.be/2rf97E2fl 14. 15. McLaughlin, Casey. 2021. "Making Sampling Data Accessible." U.S. EPA, Washington, DC. 16. McComb, Martin. 2020. "Full Data Management Lifecyle - 2019, Leveraging GIS Enterprise & EPA Platforms for Emergency Response." U.S. EPA, Washington, D.C. 17. Rodgers, M., A. Speciale, T. Boe, J. Falik, and E. Silvestri. 2021. "Tools Used for Visualizing Sampling and Analysis Data During Response to a Contamination Incident." EPA/600/R-21/150. U.S. EPA, Washington, D.C. https://cfpub.epa.eov/si/si public record Report.cfm?dirEntryId=353479&Lab=CESER. 24 ------- |