Use of Vehicle Probe and Cellular GPS Data by State Departments of Transportation (2021)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

13 A brief survey was administered to a targeted audience of DOT leadership in all 50 states and the District of Columbia through AASHTO. The survey included 11 questions about state DOT use of vehicle probe and cellular GPS data for monitoring and planning activities. The goal was to determine which agencies are currently using GPS-based data, and if so, what type of data, for what purpose, and how. The complete survey questionnaire can be found in Appendix A, and the detailed survey results for each responding state are in Appendix B. This chapter summarizes the state responses to the survey. Responses to the Survey The survey was administered to 51 agencies (all state DOTs and the District of Columbia). The response rate was 90%. Forty-five states and the District of Columbia responded to the survey as shown in Figure 1. The five states in white did not respond. All respondents (N = 46) answered every question. State Use: Current and Planned The survey revealed that at least 88% of the responding states are using some form of probe- based speed data with only a single respondent not planning to use probe-based speed data in the foreseeable future. Approximately 53% of the agencies are currently using Origin-Destination data with another 22% of states planning to use it in the near future. Trajectory (breadcrumb) data and LBS had less current and planned usage with just 50% of states currently using or planning to use these data. No agencies identified new data sets not already listed in Table 1. Several agencies said they were using Waze event/incident data, but that particular data set was not deemed relevant to this study as it does not fit the definition of cellular probe/GPS data as defined. Sources of Data The predominant GPS-based data set being used by agencies today is the NPMRDS, which is made freely available to agencies and metropolitan planning organizations (MPOs) through a FHWA program. This data set is used by all 50 states and the District of Columbia as part of the Federal MAP-21 systems performance reporting requirements. Slightly fewer agencies reported procuring data directly from a supplier as shown in Figure 2. C H A P T E R 3 State of the Practice

14 Use of Vehicle Probe and Cellular GPS Data by State Departments of Transportation Figure 1. The five states that did not respond to the survey are in white above, and include Alaska, California, South Dakota, Wyoming, and West Virginia. All other states shaded in blue (and the District of Columbia) responded to the survey. Currently Use/Purchase Planning to Use/Purchase in the Near Future Do Not Use/Purchase and Have No Plans to Do So. Speed data (real-time or historic avg. speed on one or more road segments) 46 4 1 Origin-Destination data (data that describes the origin and destination of a trip) 26 11 9 Route/Trajectory data (data that describes the path and routes of a trip from an origin to a destination) 10 13 23 Location-Based Services data (data obtained from third-party apps on smartphones that periodically report the location of users) 20 5 20 Table 1. Summary of states currently using or planning to use GPS-sourced probe data. 40 34 27 25 23 13 10 10 8 6 1 1 1 0 5 10 15 20 25 30 35 40 45 Number of Respondents Using Data Figure 2. Histogram of where state DOTs purchase or acquire data.

State of the Practice 15 Use Purpose Agency usage purpose was divided into eight high-level categories as shown in Figure 3. Probe data are still not leveraged highly for responding to customer or media requests, maintenance or asset management purposes. However, 50%–80% of all agencies are using the data for planning, performance management, operations, or research. Agencies further identified more than 30 more specific application areas listed in Table 2 in descending order of frequency. Barriers to Use While the vast majority of agencies are using some form of probe data, a few agencies still have problems acquiring or using the data due to concerns over quality and/or accuracy, lack of capacity to work with large data sets, or concerns over privacy, as shown in Table 3. Some agencies noted that the penetration rates of data have increased, which is leading to higher quality data. However, this increase in penetration rate can also make the data orders of magnitude larger and more complex to work with. Working with Probe Data As shown in Figure 4, 80% of all respondents access, analyze, and otherwise work with probe- based data through a third-party analytics platform like iPeMS, the Regional Integrated Trans- portation Information System (RITIS), or StreetLight. Many also pay consultants to process and analyze the probe data on their behalf, even if they have access to an analytics platform. Consul- tants interviewed often reported using third-party analytics apps as well. Only 40% of respondents are doing some form of in-house processing and analysis with their own agency staff. Table 4 shows a breakdown of the types of third-party applications or in-house tools that are being leveraged by agencies to analyze data. The most predominant in-house tool is Excel. Far fewer agencies are using newer development platforms or coding languages to develop their own internal capabilities. 42 41 35 26 14 14 9 3 1 0 5 10 15 20 25 30 35 40 45 Number of Responses Figure 3. Histogram of the areas in which DOTs are using data.

Purpose Number of Responses Corridor Study 34 Reliability Measurement 31 Bottleneck Ranking 30 Traveler Information 30 Before-and-after Studies 29 Systems Performance Reporting 27 Travel Time Studies 27 Model Calibration 26 Travel Times on DMS 26 Project Study 24 Situational Awareness (evacuation monitoring, winter weather operations, daily operations, etc.) 24 After Action Incident Reviews 22 Origin-Destination Analytics 22 Real-Time Work Zone Management 21 Arterial Performance Measures 20 Project Prioritization/Scoring 20 User Delay Cost Analysis 20 Event Detection 19 Significant Event Analysis (bridge collapse, commuter train derailment, eclipse traffic impacts, etc.) 19 Freight Movement Analysis 18 Operations Benefit-Cost Analysis 15 Project Selection 15 Volume Estimation 14 Work Zone Impact/Audits 14 Automatically Attributing Congestion to Various Sources (signals, weather, construction, events, geometrics, etc.) 13 Responding to Citizen 13 Purpose Number of Responses Complaints Holiday Travel Forecasts 12 Project Assessment Summary 12 Signal Analytics and Retiming 12 Justifying Operations Funding 10 Public Participation Meetings 10 Trip Analytics 10 Updating the CMP 10 Select-Link Analysis 9 What-If Scenario Analysis (for development, counter- measures, lane closures, CAVs, etc.) 9 Dangerous Slowdown Detection 8 Media Engagement 8 ATSPM (Automated Traffic Signal Performance Measures) 7 Turning Movement Analysis 7 Analysis of Quick Clearance Legislation and Towing and Recovery Practices 6 Measuring Economic Impacts (of projects, information campaigns, etc.) 5 Arrival on Red/Green Analysis 4 Multi-Modal Trip Analysis (short or long-duration trips) 4 Other (please explain) 4 All Others 3 Adherence to Evacuation Notices, and Monitoring of Movements 2 Justifying Transit Spending 2 Table 2. For what purposes do agencies use probe data? Value Responses Not applicable 27 Concerns about quality and/or accuracy 8 Cost/lack of budget 7 The data are too big and/or we don’t know how to handle it. 5 Too much effort required to integrate data 4 Don’t know/understand/see the benefit(s) 2 Lack of control over data collection process 2 Other - (Please explain) 2 Don’t have a need 1 Lack of leadership support 1 Perceived concerns over privacy 1 Table 3. Why some agencies do not use probe data.

State of the Practice 17 Data Acquisition and Sharing As shown in Figure 5, approximately 40% of agencies are acquiring probe data through tradi- tional RFPs. Others have been able to justify sole source procurements because of specialized data needs or lack of competition. Other agencies purchase data through academic institution partnerships, intergovernmental agreements, or through consultant agreements. More than 20% of respondents still procure probe data through a larger procurement effort, such as a component to an advanced traffic management system or traveler information system build-out or procurement. For agencies that do procure data, some are restricted on acceptable use and data sharing, as shown in Figure 6. For example, state DOTs are not always allowed to share the data they purchase with their partners—including university researchers, first responders, other DOTs, DOT partners, and other public and private parties. More than 20% of the respondents are not allowed to share data with consultants working on behalf of the agency and MPOs. 0 5 10 15 20 25 30 35 40 45 Access data through a hosted analytics or visualization platform like iPeMS, RITIS, StreetLight, etc. Pay a consultant to analyze the data for you Process and analyze raw data directly (using in-house staff) Not applicable Number of Responses Figure 4. How agencies work with probe data. Tools that Agencies Say They Are Using Number of Agencies Reporting their use of Each Tool RITIS 29 Excel 24 StreetLight 13 ESRI or Similar GIS Tool 9 iPeMS 8 Tableau 8 R 7 Power BI 5 MS Access 4 SAS 4 Python 4 In-house Staff/Tools 3 Hadoop 2 MatLab 2 Oracle 2 SPATEL 2 Tools that Agencies Say They Are Using Number of Agencies Reporting their use of Each Tool TransCAD 2 AlgoReports 1 ATSPM 1 AVAIL 1 C++ 1 In-house Dashboards 1 Java 1 JMP 1 Lonestar 1 MS Azure Analysis Services 1 SPSS 1 SQL Server Mgmgt Studio 1 TPM PFS Task 4 1 VISSIM 1 Table 4. Tools agencies are using to analyze data (both in-house and purchased).

18 Use of Vehicle Probe and Cellular GPS Data by State Departments of Transportation 1 7 7 8 10 12 12 13 14 14 20 0 5 10 15 20 25 Not applicable Shared by a partner agency Subscription Other - Write In Through an on-call contractor Lumped together with the procurement or Operations & Maintenance of an ATMS, Traveler Information System, or other system purchase Through a non-profit agency or academic institution As part of a larger Corridor Study or Project Sole source procurement Inter-Governmental Agreement with a University or other State Through issuing an RFP Figure 5. The top 10 ways in which respondents procure third-party probe data. 1 34 32 23 17 12 11 5 3 2 0 5 10 15 20 25 30 35 40 Number of Responses Figure 6. Number of state DOT respondents who are allowed to share their probe data with different types of agency partners.

State of the Practice 19 Enabling Decision Making Ninety percent of respondents were purchasing data for use in some sort of decision-making exercise—whether that is as an input to a decision-making system or as the sole decision-making tool. Several agencies noted that the data had empowered them to make key decisions that would have otherwise been difficult or impossible to make—at least at the same level of certainty. Chap- ter 4 will provide more detailed use case examples of studies and decisions made with the probe data. Other examples provided by survey respondents are paraphrased or quoted below: • One agency decided a new bridge over the Mississippi River was feasible. • Several agencies use probe-based data for determining when it is acceptable to issue lane- closure permits based on travel demand and congestion impacts. • Multiple agencies noted that they analyzed probe data to directly inform their personnel schedules and operations center hours of operation based on travel demand: – “The operating hours of one of our district traffic management centers were extended when analysis determined that their original hours did not cover the time of day when most heavy congestion crashes occurred.” – “Holiday travel patterns tracked to help determine appropriate staffing levels.” • Many agencies noted that they leverage probe data for project prioritization: – “Probe data contributes to annual project prioritization in our “Safety & Capacity” program, but other mobility analyses are done first (such as travel demand model scenario compari- sons) and only in cases where the analysis isn’t a good fit for travel demand modeling would we then use probe data.” – “[Probe] data has been used in cases where we are trying to decide if further study is needed for a corridor study or the like.” – “Probe data is one component of our project prioritization calculation.” – It shapes our project selection, which can vary by district. – We use data to determine bottleneck locations and develop projects. – Safety Improvement Program - uses speed/congestion data to determine segments along corridors that may proactively indicate a problem. Develops projects accordingly. – Development of Micro-Simulation Standards and Component to Data Driven Project Prioritization. – Most of the probe data sources that are used by various departments within our organization are relatively new, and as such have not been fully integrated into the decision-making and project evaluation process. Often they are used to either supplement traditional sources of analysis or to evaluate the implementation of projects post hoc. One exception to this was a recent exercise to analyze a few mobility projects in our metro area that would be high-priority projects in the event that additional funding became available. In this case, the StreetLight data platform was used to estimate likely trip origins for potential users of the improved roadway and how these are distributed spatially, especially between the metro area and the rest of the state. This was done to get a rough estimate of the incidence of benefits from a project. • Signal-Related Decisions: – Signal Operations Programs - tracks regionally significant corridor performance and allo- cates resources for signal re-timing and/or fine tuning accordingly. – Changed our approach to signal system re-timing prioritization process. Use probe data instead of contracting speed studies for certain analyses. – Used mostly in before/after signal timing efforts. • Complete before-and-after studies to determine success. • Transportation Systems Management and Operations (TSMO) Strategy Evaluation: – Analyzing the effectiveness of various operational strategies and decision to appropriate funds accordingly.

20 Use of Vehicle Probe and Cellular GPS Data by State Departments of Transportation – Our TSMO Capital program funds projects largely on this data using an internally devel- oped tool. We use probe/GPS speed data to determine the congestion measures (travel time performance and bottleneck ranking) and combine them with other safety & operations data sets to rank corridors based on operational performance. Several corridors are then selected for further review and study to apply for funding for TSMO countermeasures. • Calibrated the agency Road User Cost Analysis methodology; updated the network congestion screening methodology; major shift in emphasizing operations planning. • Work Zone Management: – Deployment of smart work zone technology to automatically post queue protection messages. A traffic alerts application was developed to present probe and other crowd-sourced data to traffic center operators for improved situational awareness. • Congestion Analysis: – Statewide Model development and calibration. Congestion management process develop- ment, monitoring, and assessment of regional congestion and reliability. Project scoring and prioritize and selection funding allocations. – Identification of congestion hot spots and development of countermeasures. • Look to expand in upcoming larger corridor and area-wide plans, Project Selection, Project Prioritization, and Maintenance of Traffic. Summary of Survey Results The majority of agencies (88%) are leveraging some form of probe-based data. Speed data are the most common data form. LBS data that cover trip routes and origins and destinations are still just beginning to be used by agencies. Those who are using these data sets are largely leveraging third-party tools and technologies for their analysis. Performance measures reporting and project prioritization remain the largest uses of speed data. Trips-based data, while used less frequently than speed data, are being adopted more quickly than speed data was adopted after first introduction. Twenty-two percent of agencies are planning to use it in the near future, which would bring the total usage up to more than 75%. Many agencies are already making significant investments and operational decisions based on this relatively new data.