Analyzing Data for Measuring Transportation Performance by State DOTs and MPOs (2019)

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.

15 This chapter summarizes the state of the practice for State DOTs and MPOs in collecting and analyzing performance data, collaborating with partners, and setting targets. The chapter reflects the literature review (see Appendix A), the survey results (see Appendix B), and the case examples (see Appendix C). The survey was organized by performance area (i.e., bridge condition, pavement condition, safety, and multimodal mobility/air quality) and then by topic (i.e., collection, analysis, reporting, and application). For each topic, the survey explored current versus planned methods, tools, and gaps. The survey included • Collection—Which data is collected? How is it collected? Who collects it? What tools and technology are used? • Analysis—How is data processed, reviewed, and managed; and by whom? • Reporting—How are reports generated? How is visualization used? • Application—What methods of target setting, modeling, and forecasting are used? The results are organized according to the headings of data collection, data analysis and tools, and target setting. Bridge condition, pavement condition, safety, and mobility are all reported separately under these headings. Summaries related to collaboration, proficiency of staff, and resources are combined for all performance areas and are included near the end of this chapter. “state” and “DOT” are often used interchangeably throughout this report. Data Collection The following material describes (by State DOT and MPO) whether or not the agency has the data necessary to support each performance measure. The figures in the sections (Figures 3-1 through 3-4) show the number of DOTs and MPOs answering “yes” or “no” to the survey ques- tion: “Do you have the data item necessary to support the measure?” The figures are grouped by types of measures in the following order: bridge, pavement, mobility, and safety. In general, State DOTs have most of the data they need to report on measures. The most complete data sets are bridge and mobility, closely followed by safety, and then pavement. Bridge Data Collection—Current State of Practice Summary All of the DOTs that responded to the Bridge Condition portion of the survey (45 DOTs) have all the data necessary to report on all of the bridge condition measures. For all except “condition of culverts,” approximately one-third of the MPOs reporting (14 MPOs) do not have the data C H A P T E R 3 Current State of the Practice

Figure 3-1. DOTs and MPOs with data to support bridge condition measures. Figure 3-2. DOTs and MPOs with data to support pavement condition measures.

Current State of the Practice 17 to support the bridge measures. For “condition of culverts,” more than one-half of the MPOs responding do not have the data. The MPOs are likely relying on State DOTs for the bridge condition data. Sources of Data Most of the states (44) reported that they collect the data for the bridge measures internally. However, 18 are also collecting with contractors, and 12 are also collecting from other agencies. Open-text responses regarding sources of data included the University Bridge Inspection Program and city/county staff. Coordination For all of the bridge condition measures, almost half of the DOTs responded that they are coordinating with MPOs and local governments regarding collection. Most of the MPOs report that they are coordinating with State DOTs. MPOs reported no major issues related to bridge condition data collection. They all stated that they are working closely with their State DOTs. Pavement Data Collection—Current State of Practice Summary Forty-four DOTs responded to the Pavement section of the survey. In general, most states have the data they need for pavement condition reporting. The number of DOTs with all data to support pavement measures varies from 44 for IRI on the NHS Interstate to 34 for cracking on NHS non-Interstate. In all cases, the number of states having all data for each of the measures is slightly less on the non-Interstate NHS when compared to the Interstate NHS. Figure 3-3. DOTs and MPOs with data to support mobility measures. Figure 3-4. DOTs and MPOs with data to support safety measures.

18 Analyzing Data for Measuring Transportation Performance by State DOTs and MPOs One state indicated that it does not have the data for IRI on the NHS non-Interstate. One state also indicated that it does not have the data for rutting on the NHS Interstate. Two states do not have data for rutting on NHS non-Interstate. Four states do not have data to support faulting on NHS Interstate or non-Interstate. Seven states do not have data for cracking on NHS Interstate, and eight do not have it on NHS non-Interstate. (In answering the survey, DOTs might have been considering the requirement to collect the pavement data in both directions; however, the final rule only requires the data in one direction.) Thirteen MPOs reported for pavement and most do not have all the data required. The num- ber of MPOs having the required data varied from seven (for IRI on NHS Interstate) to five for both cracking on NHS non-Interstate and faulting on NHS non-Interstate. Generally, slightly less than one-half of the MPOs report having the required data. Sources of Data Approximately one-half of the DOTs indicated that they collect the data internally, and the other one-half uses a contractor. All the MPOs reporting collect pavement data through another agency. Coordination For all of the pavement measures, approximately 40 of the DOTs responded that they are not coordinating with MPOs regarding collection. However, 10 to 11 MPOs are coordinating with DOTs. Mobility Data Collection—Current State of Practice Summary Forty-three DOTs and fourteen MPOs responded to the Mobility portion of the survey. All 43 DOTs have the annual VMT data to support mobility measures, while 10 of the MPOs reported that they have it. Thirty-one DOTs plan to report on Posted Speed Limit (PSL). Those states who will not do so cite the following as reasons: extra work, PSL is already included in the HPMS and therefore does not need to be collected, and FHWA has promised a speed limit file. Eight MPOs reported that they will report on mobility measures separately from their State DOTs. Regarding local vehicle occupancy, thirty-three states and five MPOs indicate that they do not have the data to support the measure. However, FHWA will supply this data. State DOTs and MPOs were asked about their plans to report on multimodal measures. Tables 3-1 and 3-2 show which measures are reported by State DOTs and MPOs, respectively. Being reported? If no, do you plan to in the next 5 years? If yes, to whom are they being reported? YesMode No Yes No Internal External Both Bicycle 29.0% 71.1% 51.9% 48.2% 21.4% 7.1% 71.4% Pedestrian 23.7% 76.3% 44.4% 55.6% 16.7% 8.3% 75.0% Transit Use 61.1% 38.9% 66.7% 33.3% 10.5% 5.3% 84.2% Table 3-1. Multimodal measures reported by State DOTs (% of states reporting). Being reported? If no, do you plan to in the next 5 years? If yes, to whom are they being reported? YesMode No Yes No Internal External Both Bicycle 6 7 5 3 1 0 6 Pedestrian 5 7 5 1 1 0 5 Transit Use 9 3 2 1 1 0 8 Table 3-2. Multimodal measures reported by MPOs (number).

Current State of the Practice 19 Transit measures are the most frequently reported by State DOTs (61%). Although only 29% and 24% are reporting on bicycle and pedestrian measures, respectively, 52% and 44% plan to do so. All three measures are mostly reported to both internal and external audiences. Transit measures are also most frequently reported by MPOs (when compared to bicycle and pedestrian). All measures are also reported externally. Sources of Data Regarding VMT, all 46 responding State DOTs have the data needed to support the measure. Forty-three DOTs responded that they collect VMT in house; however, an additional 15 also responded that they use a contractor. All states except for one will be using the NPMRDS provided by FHWA to obtain speed. One state is still discussing it. For example, Pennsylvania DOT will use the NPMRDS for MAP-21/FAST Act and will use INRIX data for other metrics. Eleven out of thirteen MPOs will use NPMRDS. Thirty-eight percent of MPOs will report on PSL. However, they all reported that they are not responsible for this item. Coordination Fifty-eight percent of DOTs are coordinating with their MPOs on the topic of speed data collection. Nine states responded that they do not intend to coordinate with MPOs on the topic. Seven out of thirteen MPOs are coordinating with DOTs on speed data. Only two states do not intend to coordinate. Safety Data Collection—Current State of Practice Summary Forty-two State DOTs responded to the Safety portion of the survey and 14 MPOs did. In general, DOTs have most of the data required for the four safety measures. MPOs have less data. For Fatalities on All Public Roads, only one state out of the 42 responding does not have the data. Three out of 14 MPOs reporting do not have the data. Regarding Serious Injuries on All Public Roads, four states and five MPOs do not have the data. Only two states and five MPOs report that they do not have the data to support Rate of Serious Injuries and Fatalities per 100 Million VMT on All Public Roads. Most are collecting the data in house. Six State DOTs and five MPOs do not have the data to support Number of Nonmotorized Fatalities and Nonmotorized Serious Injuries on All Public Roads. However, all State DOTs are prepared to report on the safety measures. Sources of Data One-half of the states obtain the data in house, and the other half obtain the data from other agencies. Other sources cited include NHTSA’s FARS, law enforcement, highway patrol or police, and Departments of Motor Vehicles (DMVs). Thirty-eight responses were received to the question “How do you plan to handle all public roads (crash)?” They all indicate that crash data is collected on all roads, and FARS maintains fatal crash information on all public roads. Detailed responses can be found in Appendix B in the section on State safety measures. When asked how they plan to handle VMT on all public roads, some responded that they are collecting traffic on all public roads; however, most indicate that they do not have the vol- ume data readily available. DOTs stated they would obtain VMT by collaborating with local

20 Analyzing Data for Measuring Transportation Performance by State DOTs and MPOs governments, from HPMS, or use estimation and sampling techniques. For example, one state responded, “Counts are taken on representative samples of all roadway classifications, with uncounted locations estimated based on these samples and the statewide network of automatic traffic recorders.” Another state responded “The DOT through the normal processes and in response to the Minimum Inventory Roadway Elements (MIRE) Fundamental Data Elements (FDE) are working with locals to ensure we will have the VMT information to ensure we can calculate the rate measures appropriately for all public roads.” Coordination More than half of responding States indicated that they are collaborating with local govern- ment to obtain the data. All of the responding MPOs are collaborating with State DOTs. State DOTs generally have the data they need to support MAP-21/FAST Act measures; however, they do have some concerns about the data. Data Analysis and Reporting Tools Most states are using specialized tools and technology to analyze and report their data. Figures 3-5 through 3-8 show what percentages of State DOTs are using special tools for analysis and reporting. Specialized tools and data are used most in the bridge and mobility performance areas (29 for bridges and 27 for mobility). The next highest use of specialized tools occurs in the safety area (24 states). Only 11 states use specialized tools in the pavement area. Appendix A includes descriptions of some of the main tools used by states. Figure 3-5. DOTs using specialized technology/tools to analyze data. Figure 3-6. DOTs using specialized tools or visualization methods for reporting on performance measures.

Current State of the Practice 21 Bridge Data Analysis Twenty-nine State DOTs are using specialized technology and tools. The specialized tools include AASHTO BridgeWare Management Software (11), in-house database tools (6), Bridge Condition Forecasting System, Deighton dTIMS (3), Maine DOT Business Intelligence, state- specific tools such as Minnesota DOT’s Bridge Replacement and Improvement Management (BRIM) and Tennessee Roadway Information Management System (TRIMS). Wyoming DOT uses historical data and has developed its own deterioration models, improvement models, and cost models, in addition to optimization algorithms. Pavement Data Analysis Of the 44 responding State DOTs, 32 are using specialized technology and tools to analyze pavement condition data. They include DTIMS (8 DOTs), in-house (10), and others (WiseCracks, AgileAssets, ARAN, Stantec, and Fugro). Mobility Data Analysis Twenty-seven of the states reporting are using specialized tools. They include CATT Lab (RITIS/VPP) (9), GIS (5), INRIX (2), Performance Measurement System (PeMs) (2), Tableau (3), Figure 3-7. State DOTs with a process for target setting. 26% Figure 3-8. Extent of collaboration with MPO partners as reported by DOTs.

22 Analyzing Data for Measuring Transportation Performance by State DOTs and MPOs travel demand models (5), MATLAB, SAS and R, SPATEL from TRANSCOM, and PostgreSQL. Two notable practices follow: • Pennsylvania DOT is undertaking several efforts internally and with universities to calculate performance measures using custom-developed tools. These efforts are focused on joining several data sources together to create a picture of congestion/mobility. • Wisconsin DOT does not have the technology/tools to analyze the mobility data. The UW TOPS Lab mentioned earlier has the technology, tools, and ability to do the analysis. Safety Data Analysis Twenty-four of the states reporting are using specialized technology and/or tools to analyze safety data. These tools and technology include AASHTOWare SafetyAnalyst, Excel, SAS, GIS, Tennessee Integrated Traffic Analysis Network (TITAN), Tableau, Cognos, SPSS and JMProm Highway Safety Manual, and GEARS (Georgia tool) and in-house tools such as Safety Voyager (a product developed by New Jersey DOT) and the Highway Safety Manual (HSM) predictive method/SafetyAnalyst, in conjunction with Virginia-specific SPFs. As noted in the Case Example for NOACA, NOACA uses a tool to perform sliding-scale corridor analysis. More DOTS are using specialized tools and visualization methods for reporting on performance in the pavement area than any other area. Mobility tools for reporting are used in 22 states, followed by safety tools in 18 states and bridge tools in 18 states. DOTs reported using tools to analyze data in the bridge area at a much higher rate (29 states) than they use tools to report for bridge condition (18 states). DOTs use specialized tools for reporting in the mobility and safety performance areas, and slightly more for data analysis than for reporting on performance. Bridge Data Reporting Tools Eighteen states are using specialized tools or visualization methods for reporting. These tools and methods include AASHTOWare BrM, Bentley InspectTech, Crystal Reports, Deighton dTIMS, Excel and GIS, InDesign, Pontis, NBI, and an in-house spreadsheet to determine future condition based on an NBI scale. Other tools, reported by MPOs, include DecisionLens and NBIAS. North Central Texas Council of Governments (NCTCOG) reported PONTEX is the current bridge inspection data-management system employed by the Texas DOT. It features a web-based, role-based relational database structure designed to replace the BRINSAP legacy mainframe system, which had been in service for decades. First implemented in 2009 and now utilized by all Texas DOT districts, PONTEX makes all bridge inspection and inventory data readily available to all end users throughout its own various applications, or through a link to the relational database used to store the data. PONTEX is intended to be a major data source incorporated into “Decision Lens,” a relational performance-based, decision-making tool aimed to guide project prioritization and program- ming for the Texas DOT Unified Transportation Program (UTP). NCTCOG has recently been granted authorization to access and manipulate project information within the “Decision Lens” tool, but it is unknown whether or not NCTCOG will also be able to access specific bridge data via PONTEX, or what influence the agency will have on state actions according to bridge or other data highlighting North Central Texas projects. Pavement Data Reporting Tools Twenty-four of the states that responded to the Pavement Condition portion of the survey are using specialized tools or visualization for reporting. These tools and methods include

Current State of the Practice 23 ArcGIS, dTIMS, Excel, InDesign, Tableau, Oracle Business Intelligence, Agile Assets Pavement Analyst reports, and COGNOS. New York State DOT uses an internal pavement data viewer that allows for synchronized viewing of location maps, right-of-way video logs, and pavement images, along with certain pavement data. NOACA uses RoadMatrix Pavement Management software, and NCTCOG uses North Texas Share Pavement Analysis Services. Mobility Data Reporting Tools Twenty-two states are using specialized tools for reporting on performance. These tools include many of the same tools as reported for the previous topic related to data analysis: RITIS, GIS, ITERIS, PeMS, and Tableau. Some exemplary practices are described below: • New York State DOT, in collaboration with the state’s 14 MPOs, working through its Univer- sity Transportation Research Center (UTRC) and using the State University of New York’s AVAIL, has analyzed the NPMRDS data and developed a tool that will be used to produce the required metrics. • RITIS, developed by the University of Maryland, is a popular tool for both mobility data analysis and reporting. State DOTs can use AVAIL and RITIS to analyze data and develop reports. More information on AVAIL and RITIS is available in Appendix C. Safety Data Reporting Tools Eighteen DOTs responded that they are using specialized tools or visualization methods for reporting on safety performance measures. These tools and methods include many of the same ones reported for the data analysis question: Excel, GIS, and Tableau. Some exemplary practices are described below: • Texas DOT uses Microstrategy Business Intelligence Tool to extract and analyze crash data and Microsoft Excel for projections purposes. • Washington State uses performance reporting through the Grey notebook. Washington State DOT has developed a data portal for local transportation agencies. The DOT is also incorpo- rating the portal into SafetyAnalyst to use for planning. • Wisconsin DOT uses a dashboard tool, Xcelsius by SAP, to display and report safety and other departmental measures. A PDF report is also used for reporting, which is developed using Adobe’s InDesign. WisDOT also has a tool—Community Maps—that provides Wisconsin’s local law enforcement and county Traffic Safety Commissions with an online interface for mapping crash data. Crashes are mapped through a combination of manual and automated processing. Target Setting Figure 3-7 shows how many State DOTs have developed a process for target setting for each of the four performance areas. (CMAQ was not included in this assessment.) State DOTs are most advanced with respect to developing target-setting methods for safety measures with 40 DOTs reporting they have methods in place. Twenty-eight DOTs report they have methods in place for bridge target-setting and 22 states have methods in place for pavement target-setting. Only 13 states have developed methods for mobility target setting.

24 Analyzing Data for Measuring Transportation Performance by State DOTs and MPOs Bridge Target Setting Twenty-eight states have developed bridge target-setting processes. Most are based on historic trends, models into the future or goals. Most define their targets as realistic (i.e., in line with plans, but doable per the survey, as opposed to the other two choices (i.e., minimum or conserva- tive (easy to meet) and stretch or aspirational). Only 2 of 13 MPOs responded that a process for target setting had been developed. All respondents answered the question regarding their plans for target setting. Eleven indi- cated that the process would be based on historic trends, 11 indicated they would be forecast, five indicated they are in process, and the rest were using detailed processes as described below: • Michigan: “We have been forecasting bridge condition using our Bridge Condition Forecast Spreadsheet for many years. This spreadsheet takes into account current condition, historic deterioration rates, project costs, future budgets, and currently programmed projects in order to project future network-level condition based on a minimum NBI condition rating. We intend to continue with this spreadsheet to develop targets for bridge condition with perhaps a more detailed analysis on programmed projects to evaluate open to traffic dates in order to better determine when those projects would tie to a final inspection and an increase in condition.” • California Department of Transportation (Caltrans): “California was required to develop con- dition targets for bridges under existing state laws. These targets were developed considering different funding and performance scenarios. The targets adopted by the California Transpor- tation Commission reflect a condition that is better than the current condition. Targets that were developed and approved only pertain to the State Highway System (SHS) bridges, and include many bridges owned by the state that cross over the SHS. Many of these bridges would not be considered part of the NHS, but are the responsibility of Caltrans to manage. Caltrans is responsible for more than 90% of the NHS bridge deck area in California, and most likely the NHS targets set for MAP-21/FAST will mirror this ownership breakdown.” • Wyoming: “Uses an in-house developed optimization algorithm to predict the effect of various budget scenarios on bridge conditions. Tradeoff analysis is performed between budgets and achievable bridge conditions; and based on these results, a realistic target is developed.” • Oregon DOT: “1) Prepare tables and graphs of the trend line for the previous 5 years; 2) Make projections of the trend line into the future; 3) Adjust the projected trend line based on pro- jected funding level, agency goals, and projected deterioration using agency-developed models; 4) Collaborate with stakeholders and other program target-setting efforts; 5) Prepare a report that summarizes the background and approach for proposing draft targets; 6) Present the report and proposed targets to decision-makers in Oregon DOT Management and the Transportation Commission who will establish the final agency targets.” • NOACA: “Regarding plans for target setting for MPOs, MPOs plan to coordinate with their states. NOACA is the only one who has set targets and aligned them with their strategic plan, and has benchmarked with Oregon DOT’s targets.” Seventy percent of states indicated that they would use tools to forecast future values of per- formance. Tools cited include AASHTO BrM (8), Deighton dTIMs (5), in-house Access data- bases (7), and other in-house tools (11). Pavement Target Setting Twenty-two of the DOTs have set a process for target setting for pavement. Most of these processes are based on historic trends, models, and goals. Seventy percent indicated that the targets will be realistic (i.e., in line with plans, but doable), as opposed to the other two choices (i.e., minimum or conservative (easy to meet) and stretch or aspirational). Only 2 out of 12 MPOs responded that they have set targets in the pavement area.

Current State of the Practice 25 Some practices by DOTs related to pavement target setting are as follows: • Michigan DOT: “For the First Performance Period, Michigan DOT plans to collect and ana- lyze data, and then use historical trend analysis to develop targets. Michigan DOT’s targets and method will require the Director’s approval before final coordination with the MPOs. The development of this process is in progress. In the longer term, Michigan DOT will develop or seek new forecasting tools based on the new metrics to aid in target development.” • Ohio: “Already has goals and trends on its own Pavement conditions and ratings; this will be redundant and different from Ohio’s goals. Ohio has been using our pavement goals and rating for over 25 years.” • New York: “Program area managers will collaborate in the near future to discuss and develop pavement condition targets. Once the State has developed draft targets, these will be shared and discussed with the MPO’s.” • Caltrans: “Caltrans will use performance cost curves based on expected deterioration and average unit costs to provide a range of performance options and associated costs. Using these curves and historical condition levels, we will make a determination of an appropriate target . . .” Thirty-five State DOTs use tools to forecast future values. They include DTIMS (10 states), HPMA (Stantec) (1 state), in-house (13 states), and Agile Assets (3 states). Only one state does not plan to use models to predict pavement targets in the future. Mobility Target Setting Thirteen states have developed a process for mobility target setting. Most processes are based on historic trends, goals, and consensus with stakeholders. Twenty-four of the 34 states respond- ing to the question indicated that their targets are “realistic” (i.e., in line with plans, but doable), as opposed to the other two choices (i.e., minimum or conservative (easy to meet) and stretch or aspirational). Only one MPO reported that it has set mobility targets. Regarding plans for developing targets, 36 responses were received. Ten states have not developed plans yet and are waiting on further direction. Several are basing targets on historic trends. Other plans for developing targets include developing a baseline once data is available, coordinating with MPOs, using RITIS, using NPMRDS, relying on consultants, and taking advantage of safety workshop target setting. Other methods are explained in the following quotes from the survey: • Wyoming: “We started with the output of the last 3 years of data to find our own baseline. We have been running scenarios of target setting to get indications of failure points. We brief the MPO technical working groups on the findings, and then the Executives at Wyoming DOT. Upon approval of the targets by the executive staff, the MPO policy boards are briefed.” • Florida: “Florida DOT and the state MPOs have virtually agreed on an approach and now are waiting for the data to finalize the targets. We will be meeting for probable final concurrence within 6 months.” • Maryland DOT SHA: “We have long produced a Mobility Report where we have looked at mobility in a number of ways and provided forecasts. We are looking at baseline data now and developing trends and forecasts. We plan to meet with the MPOs to discuss this and to coordinate further for target setting. We plan to use baseline data and the forecasts to develop realistic targets and to consider where land use is changing, population shifts, economic devel- opment, etc. in our target setting. We will also think about planned projects and operational strategies that will improve congestion so that we can identify the best target. For example, we may set a more aggressive target if we can identify a good Transportation System Management and Operations (TSM&O) approach that we think will work more quickly.”

26 Analyzing Data for Measuring Transportation Performance by State DOTs and MPOs • Iowa DOT: “This will be impacted by anticipated federal guidance. However, the process is likely to be similar to what was used for safety measure target setting. Generally, this involved early and frequent stakeholder discussions to outline the process, followed by a discussion and evaluation of various target-setting methods. Ultimately, the selected method was a risk- based approach that used historical data without the influence of external factors. The draft targets were then reviewed by stakeholder groups and presented to decision-makers for final approval.” • North Dakota: “Technical staff will likely prepare trend-based scenarios for internal experts to consider. Then, the internal experts add to or delete scenarios, as appropriate, resulting in ±3 scenarios on which external stakeholders comment. The final targets are set by the North Dakota DOT Director, based on all obtained input.” Sixteen states and nine MPOs use tools to forecast future values of performance for mobility. Most responded that they use a statewide or MPO Travel Demand Model. Twelve states and two MPOs do not plan to use tools to forecast mobility. Safety Target Setting Only two of the responding states have not developed their target-setting processes. Five out of 14 MPOs have developed safety targets. Most are based on historic trends; however, states also indicated use of models, goals, and consensus with stakeholders. Most states indi- cated that their approach is realistic (i.e., in line with plans, but doable), as opposed to the other two choices (i.e., minimum or conservative (easy to meet) and stretch or aspirational). Twelve MPOs responded regarding plans for target setting. Most are working closely with their State DOTs. Missouri DOT has a well-developed safety target-setting process, as documented in Appendix C. Thirty-seven responses were received from states regarding plans for setting targets. Practices reported include the following: • Wyoming DOT: “We started with the output of the last 10 years of data to plot trend lines. We then briefed internal leadership (primarily Districts) on a tentative target, then sought input from the MPOs, then requested a decision from the Executive Staff.” • Virginia: “Virginia DOT and DMV work collaboratively to set safety measure targets by considering historical trends and future projections. We’ve had a FHWA/NHTSA workshop with MPOs and other partners to talk about safety target-setting coordination.” • Arkansas State Highway and Transportation Department: “We recently had FHWA resource center come in and provide a safety target-setting workshop that included the DOT, State Safety Office, NHTSA, state motor carrier safety office, and MPOs. From this workshop, an action plan is being developed on how to set targets, and how each partner can help meet those targets.” • West Virginia: “West Virginia has adopted the goal of halving fatalities by 2030, and is making significant progress toward that goal. Targets involving fatalities are being established using historic data and projecting that forward toward the goal; however, we understand that through this process outside influencing factors, such as the economy or changes in legisla- tion, will need to be considered and targets may require adjustment. West Virginia has not yet established serious injury number/rate goals. Looking at the historic data, it appears the serious injuries are declining much more rapidly than fatalities. As such, West Virginia is considering adopting a goal of a two-third reduction by 2030. The same method, as described above, will be used for serious injury, serious injury rate, and bike/pedestrian fatality targets, although the goals are still under review.”

Current State of the Practice 27 • Rhode Island: “Rhode Island DOT has adopted an aspirational goal that results in a calculated 3.2% reduction every year.” • Maine: “DOT to be working with Maine Bureau of Highway Safety to establish targets as pre- scribed. May have supplemental targets—yet to be determined. Have been comparing actual results to our already established 2016 targets, and will use those findings to influence future target setting. Have met with MPOs and have an overall philosophy that they will establish safety targets based on statewide targets—but will be established based on the MPOs’ deter- mined historical benchmark.” • New Jersey: “New Jersey DOT has already developed targets in collaboration with MPOs and Department of Highway Traffic and Safety. We looked into the historic trend, took into account the economy and projects done in the past 3 years and reduction we may expect etc. to name the few things.” • Wisconsin: “Common practice for Wisconsin DOT has been to use an annual 5% reduction from the 5-year rolling average.” Twenty-six states and one-third of MPOs are using tools to forecast safety performance values. These tools include 5-year rolling average, Excel, SAS, IBM Watson, regression models, SafetyAnalyst, trend analysis, historical data, HSM methods, and activity-based and travel demand models. Nine states and five MPOs do not plan to use tools to forecast in the future. Most states indicated that their approach is realistic (i.e., in line with plans, but doable), as opposed to the other two choices (i.e., minimum/conservative and stretch/aspirational) for all of the performance areas. Collaboration Figure 3-8 shows the extent of collaboration with MPOs as reported by DOTs for each of the four performance areas. DOTs report having the most developed collaboration with MPOs in terms of safety mea- sures, followed in order by mobility, bridge condition, and pavement condition. Several states have well-developed coordination processes. Missouri DOT was featured in an FHWA TPM Noteworthy Practice for the process (See Appendix C). Missouri DOT also devel- oped an extensive Excel spreadsheet to document all deadlines, reports, and requirements for the MAP-21/FAST Act. Michigan DOT also has a well-developed process and has assigned a TPM Implementation Team with coordinators for each of the performance areas. A key challenge for transportation agencies today relates to maximizing the utility of the numerous data streams available to support asset management, performance management, transportation system management, and operations. Several states are developing Data Business Plans (DBPs) to assist State DOTs and local partners understand what roadway travel mobility and safety data is being collected within their organizations, how the data supports key mobility and safety initiatives, and what improvements related to data processes are necessary. A DBP plans for efficient use of people, processes, and technology; links business objectives, programs, and processes to data systems, services, and products; and guides an agency in data-management practices. More details and an example can be found in Appendix C (Case Examples). Proficiency of Staff Tables 3-3 through 3-6 show the levels of staff and consultant proficiency for each area of performance. The DOTs and MPOs were asked to rate overall proficiencies for data collection, analysis, and reporting capabilities for the items listed.

28 Analyzing Data for Measuring Transportation Performance by State DOTs and MPOs No demonstrated proficiencyCategory Some knowledge Satisfactory Somewhat proficient Fully capable and proficient Technical capability of staff 0.0% 0.0% 2.2% 11.1% 86.7% Technical capability of consultants 7.7% 2.6% 5.1% 20.5% 64.1% Current analysis tools 0.0% 4.4% 20.0% 44.4% 31.1% Table 3-3. DOT bridge condition measures proficiency ratings. No demonstrated proficiency Some knowledge Satisfactory Somewhat proficient Fully capable and proficient Technical capability of staff 0.0% 4.8% 4.8% 16.7% 73.8% Technical capability of consultants 0.0% 3.2% 9.7% 32.3% 54.8% Current analysis tools 0.0% 14.6% 9.8% 41.5% 34.1% Category Table 3-4. DOT pavement condition measures proficiency ratings. No demonstrated proficiency Some knowledge Satisfactory Somewhat proficient Fully capable and proficient Technical capability of staff 4.5% 31.8% 27.3% 20.5% 15.9% Technical capability of consultants 7.9% 18.4% 21.1% 21.1% 31.6% Current analysis tools 6.8% 27.3% 34.1% 18.2% 13.6% Category Table 3-5. DOT mobility measures proficiency ratings. No demonstrated proficiency Some knowledge Satisfactory Somewhat proficient Fully capable and proficient Technical capability of staff 0.0% 7.5% 5.0% 27.5% 60.0% Technical capability of consultants 2.9% 5.9% 17.6% 32.4% 41.2% Current analysis tools 0.0% 5.0% 32.5% 27.5% 35.0% Category Table 3-6. DOT safety measures proficiency ratings.

Current State of the Practice 29 Many DOTs reported that their staffs are fully capable and proficient with data collection, analysis, and reporting for bridge condition measures. Forty-four percent of states reported that they are somewhat proficient with the current analysis tools. Similar results are observed for the pavement condition measures. Many DOTs reported that their staffs are fully capable and proficient with data collection, analysis, and reporting for pave- ment condition measures. None of the states reported that their staff or consultants have no demonstrated proficiency. More than half of the states reported that they are somewhat profi- cient with the current analysis tools. In the case of mobility measures, DOTs report that consultant technical capability is higher than staff capability. There is also room for improvement in the proficiency with current analysis tools with approximately one-third of states reporting satisfactory proficiency. Safety is the only performance area where DOTs reported the highest percentage in the fully capable and proficiency column for all three areas (technical capability of staff, consultants, and current tools). Tables 3-7 through 3-10 show responses from MPOs. The numbers are shown in actual responses (rather than percentages). Five MPOs report that their staffs have full proficiency with bridge condition measures. No demonstrated proficiency Some knowledge Satisfactory Somewhat proficient Fully capable and proficient Technical capability of staff 3 4 1 1 5 Technical capability of consultants 3 1 0 1 3 Current analysis tools 2 4 2 3 3 Category Table 3-7. MPO bridge condition measures proficiency ratings. No demonstrated proficiency Some knowledge Satisfactory Somewhat proficient Fully capable and proficient Technical capability of staff 3 1 2 2 5 Technical capability of consultants 3 0 0 2 3 Current analysis tools 2 1 5 1 4 Category Table 3-8. MPO pavement measures proficiency ratings. No demonstrated proficiency Some knowledge Satisfactory Somewhat proficient Fully capable and proficient Technical capability of staff 2 0 1 1 8 Technical capability of consultants 1 0 1 1 2 Current analysis tools 2 0 6 1 3 Category Table 3-9. MPO mobility measures proficiency ratings.

30 Analyzing Data for Measuring Transportation Performance by State DOTs and MPOs No demonstrated proficiency Some knowledge Satisfactory Somewhat proficient Fully capable and proficient Technical capability of staff 1 0 1 8 4 Technical capability of consultants 0 1 0 4 1 Current analysis tools 1 0 5 6 2 Category Table 3-10. MPO safety measures proficiency ratings. Results for pavement measures are similar to bridge; however, there is slightly more profi- ciency with current analysis tools related to pavement. MPOs report being more proficient with mobility measures than with all of the other measures. Most MPOs report that their staff, consultants, and proficiency with current analysis tools are somewhat proficient. The survey results show that DOTs need to improve proficiency in staff, consultant, and tool capability for all performance areas with mobility having the highest need, followed by bridge, pavement, and safety. Resources State DOTs and MPOs were asked if they have sufficient resources to handle the data col- lection, analysis, and reporting of measures (within each of the four performance areas). Figure 3-9 illustrates the results. State DOTs reported that they have the most resources to handle bridge condition measures (with only one state having insufficient resources), followed in order by pavement condition, safety, and mobility. Almost one-half of the states reporting have insufficient resources to handle data to support mobility measures. Figure 3-9. DOT availability of resources to handle the data to produce the measures.