Visualization of Highway Performance Measures (2022)

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.

49   This chapter will discuss five case examples selected to provide a deeper dive into the visualiza- tion of performance measures. Case examples come from interviews with state DOT representa- tives and represent current practices from the DOTs listed in Table 4. Case example projects were selected from DOT survey respondents according to the following criteria: • A cross-section of state DOTs across the country by geography and size. • Experience reporting performance measures. • Success in developing strong programs. • Current visualization approaches. Table 4 summarizes the five case examples with a short summary statement and notes the primary visualization tool the state DOT discussed. The case examples reflect the responding state DOTs’ experiences. Washington State DOT and Virginia DOT both addressed a pressing problem that was negatively impacting their agencies by defining performance measures and using visualizations to communicate them to their target audiences. WSDOT faced a lack of trust in their agency. The perception of Virginia DOT (VDOT) was their projects were not delivered on time and cost more than expected. Because so many per- formance measures were reported on highway congestion, transportation solutions tended to be highway centric. Florida DOT helped the planning community adopt multimodal performance measures to communicate how people and goods are moved using all modes, including on high- ways. Utah DOT is rapidly growing, and, applying the lessons learned from the successes of other state DOTs, chose to build a comprehensive dashboard that tells the performance story of their efforts. Arkansas DOT closely observed the progress of other state DOTs, and seeing the benefits they gained from their visualization efforts, followed their lead to emulate the successful outcomes in the areas of most concern to their state. C H A P T E R   4 Case Examples State Experience Summary Visualization Tool WA Using performance journalism to tell agency story Gray Notebook VA Dashboard makes an impact Dashboard UT Telling the Utah performance story Dashboard FL Defining multimodal mobility performance measures Source Book AR Following the leader to lead by example Dashboard Table 4. Summary of state interview experience and primary visualization tools.

50 Visualization of Highway Performance Measures Washington State DOT: Using Performance Journalism to Tell the Agency Story Washington State DOT has pioneered accomplishments in the visualization and communica- tion of performance measures; some examples are shown in Figure 36. Nearly two decades ago, WSDOT faced a problem of credibility and accountability that was affecting the ability to attract the investment of necessary support and resources to accomplish desired goals as effectively as possible. The idea of performance management originated from the top-down as a commitment to the people of Washington State, and their legislative representatives, to restoring transparency, accountability, and credibility in their ability to meet performance expectations. Through their extensive efforts, they have been able to create and maintain a growing system of award-winning data visualization tools. When this effort began, it was pure innovation. No one had been professionally trained in visualizing performance measures, so it was necessary to envision and create the tools from scratch. One of the first and most important lessons learned in this process was the central importance of audience in directing that process. One WSDOT representative who has been involved with this program from the beginning emphasized how they learned this lesson through trial and error. Several years into this visualization project, the fledgling division was asked to present some of their visualizations to communicate performance measure data to the governor and other state-level decision-making officials. Confidence and excitement were high because the WSDOT team had up to that point created many data visualizations that had been greeted with positive feedback within the transportation community. They were excited to showcase one visualization tool in particular, a heat map that shows where particular types of activity are occurring at a given location. However, despite the success of their other visualization tools at communicating performance data successfully at this meeting, the heat map was received with confusion as to how to read it and what information it was trying to convey. WSDOT was able to benefit from this by identifying the significance of using the most appropriate visualization to effectively communicate with a clearly defined audience. It was also at this early stage of development that a WSDOT practitioner conceptualized the visualization of performance measures as “performance journalism.” This concept defines the practice by conveying that the goal and intent of it is to communicate distinct stories, in the jour- nalistic sense of explaining the who, what, when, and where of what is most currently happen- ing in the transportation system. This conceptualization of their work as a form of journalism led them to two important decisions that have contributed to their stated goals of transparency, accountability, and credibility. One was the decision to design their visualizations at the audi- ence level of the general public. This set the standard, going forward, for clarity and simplicity of design at the highest possible effective level. Another was the decision to not create a firewall between what types of data or visualizations would be available publicly versus internally. The concept of journalism invokes an ethical standard of accuracy and transparency to tell stories based on facts, regardless of whether those facts reveal roaring successes or areas in need of significant improvement. From the beginning, WSDOT believed in the power and value of this information to improve the overall performance of the state transportation system as a whole (multimodal), as well as the ways in which the state transportation system intersects with both national and inter- national systems. Therefore, the intent and purpose from the earliest years was to speak to those who could most use this information to inform decisions, particularly in terms of where to best allocate efforts and resources, and what efforts and resources would be most effective. These audiences were defined as the general public, elected officials, and other stakeholders in the transportation community. While that definition is broad, the definition is still important because it identifies that multiple tools with varying levels of specificity would be required to achieve success with each group within the overall intended audience.

Figure 36. Washington DOT visualization examples (Source: WSDOT).

52 Visualization of Highway Performance Measures The initial culmination of WSDOT’s efforts was the quarterly PDF publication of a docu- ment they titled the Gray Notebook. As of this writing, WSDOT has issued around 80 editions of the notebook. The amount of performance data visualization that is represented in this body of work is enormous, particularly when one considers that it has all been created on a learn-as- you-go basis. This is notable when considering the accomplishments of WSDOT as it invites one to consider what is possible in this field moving forward, when, perhaps, even more support and resources could be made available for innovation. While the initial support for mobiliz- ing a performance management approach through data visualization came from the top-down, once the Gray Notebook project was launched, additional support began to come from end users of the report. More and more stakeholders began requesting more stories be told, both in terms of performance measures they wanted to see visualized and of their own data stories that they also wanted to be told to the wider community. At one point, the publication grew to over 100 pages, which is when WSDOT decided to re-emphasize the concept of “performance journalism” by strictly limiting inclusion in the report to stories that could be told entirely by visualizing the data. In 2007, Washington State passed a law establishing specific policy goals for transportation agencies. This allowed WSDOT to even further streamline their publica- tions to speak to progress toward the state’s specific policy goals. The Gray Notebook currently reports on progress toward six of these statewide goals: Safety, Preservation, Mobility, Environ- ment, Economic Vitality, and Stewardship. Performance Journalism WSDOT showcased how they apply performance journalism using the different visualization tools discussed in this synthesis. Simple Charts Simple charts as shown in Figure 37 are the most dominant data visualization tool used to share data. When WSDOT is designing these visualizations, the title and subheading of each of these examples is carefully crafted to tell the story. Infographics Like the headline of a newspaper, the first page of each edition of the Gray Notebook is titled “Performance Highlights.” This page is presented as an infographic, as shown in Figure 38, that contains brief text headlines, incorporated within relevant graphics, to communicate to the reader the most significant stories contained within the document. For example, one publication uses a stylized graphic of a suspension bridge that incorporates the headline: “292 Bridges owned by WSDOT are currently over 80 years old.” The economical use of text in a headline format accompanying visualizations can be described as a standard prac- tice that is used throughout WSDOT’s visual reporting of performance measures. This practice allows for the clear and easy identification of the most important story or message that a par- ticular visualization is intended to communicate, therefore immediately alleviating the need of the user to try to “figure out what they’re looking at.” It also allows those who may not need the specifics of the data in each visualization to still get the essence of the story at a glance, again, sim- ilar to how headlines function in traditional media journalism. Another benefit of this practice that WSDOT representatives emphasized as a standard is that the use of such headlines contributes to the goal of creating visualizations that can be extracted from the document and used without the necessity of referring to the original document for clarity or explanation. The visualizations are intentionally designed to communicate the performance story by standing alone. A strong example of how all these principles come together in the Gray Notebook is a visual- ization comparing crash rates during the pandemic and the percentage of crashes that resulted

Simple Charts Bar charts were used in 83% of Appendix C visualizations. Pie charts were used in 29% of Appendix C visualizations. Line charts were used in 46% of Appendix C visualizations. Tables were used in 35% of Appendix C visualizations. Bar Chart Pie Chart Line Chart Table Figure 37. Simple chart examples (Source: WSDOT). Figure 38. Infographics examples (Source: WSDOT).

54 Visualization of Highway Performance Measures in serious injury or fatality to pre-pandemic rates (see Figure 39). The headline accompanying this visualization reads: “Total crashes involving active transportation users decline during pan- demic, but a higher percentage resulted in a fatality or serious injury compared to pre-pandemic average.” The message is simply conveyed to the casual reader immediately. For those readers who desire more detailed information, who want to see the data that support that story, it is clearly and simply presented in a chart. A simple key is provided at the top that labels the three different color-coded elements visualized in the chart and what they represent: (1) a bar graph showing total crashes over time, (2) a dotted line on the timeline that represents the beginning of the pandemic, and (3) a line chart that represents percentage of serious injury or fatality. In describing this example, WSDOT representatives explained the necessity of avoiding busy graphics whenever possible, because they have found that their audiences find them to be con- siderably more confusing. Simplicity leads to increased clarity across a wider-ranging audience. Another tool that WSDOT has found to be effective in creating visualizations for the Gray Notebook that can often be overlooked because of its simplicity and obviousness is the photo- graph. For example, in visualizing information about what is being done by WSDOT to keep its workers safe throughout the pandemic, a simple photograph of a transportation system worker on the job properly wearing personal protective equipment and practicing social distancing, Figure 39. Example of performance journalism (Source: WSDOT).

Case Examples 55   along with a brief caption, does the job. A particularly impressive visualization created to com- municate the performance of pavement maintenance incorporates the use of photographs into a larger infographic. The headline for this visualization is: “Percent of pavement in very poor condition decreases between 2015–2019,” as seen in Figure 40. Again, the overall story can be easily understood from the headline. The infographic that follows the headline then contains a chart formatted in several rows and columns. The first column is titled “What Drivers See,” and contains a column of photographs. The rows represent different stages of pavement deteriora- tion. Anyone not already familiar with stages of pavement condition, but familiar with what they see when they are driving, can immediately connect their experience with the rest of the data being presented. The next column contains graphics that illustrate what is happening at each level of the pavement (road surface, pavement base, and dirt/gravel base) at each corresponding level of deterioration. Another column then provides line graphs about lane miles and VMT, followed by another column with a brief text entry explaining simple facts about each level of condition. Finally, there are two columns representing the desired trend versus the current trend with the simple use of up or down arrows. This infographic, strategically incorporating the use of photos, allows for the presentation of data about a relatively specialized topic in a manner that most people can easily understand. WSDOT has also made use of aerial photography, which is particu- larly useful for monitoring the performance of environmental mitigation measures. Figure 40. Infographic using photography, graphics, and line charts in a table (Source: WSDOT).

56 Visualization of Highway Performance Measures Maps, as seen in Figure 41, are yet another visualization tool that WSDOT has found to be effective through the Gray Notebook. In addition to aerial photography, maps have been used to tell other stories of relative progress toward statewide environmental policy goals. For example, one Gray Notebook visualization uses a map of Washington State and its counties to commu- nicate information about the total number of registered plug-in electric vehicles by county. Counties are represented in deepening shades of green to visually communicate higher and lower numbers, with exact figures for each county displayed within their section on the map. This visualization tells the stories of each individual county’s performance in this area in a manner that is clean, simple, and concise. One question is whether the Gray Notebook project has created tools for data communica- tion that have been influential in making real-time decisions. One of the challenges of being a pioneering institution is that by the time what that institution has been doing gains traction, that institution has already been doing it for two decades. Progress, therefore, can tend to feel slow or appear relatively ineffective. However, it is important to include here that any time innova- tion is involved, there is a relative learning curve that corresponds to the introduction of that innovation. As a result, WSDOT identifies among its goals what can effectively be described as education. It is not enough to simply present or make available these tools; it is also required to educate stakeholders in the system about the value of the tools and practical ways in which Figure 41. Map example (Source: WSDOT).

Case Examples 57   they can be applied. Like any educational endeavor, over time, an increasing number of stake- holders are being exposed to the value and applicability of these tools. The Gray Notebook has produced a number of success stories in terms of its usefulness as a tool for guiding decision making. For example, there is the story of a simple, color-coded, stacked line chart designed to communicate the percentage of concrete bridge decks that were due or past due for repair between 2016 and 2020, as shown in Figure 42. This visualization makes it easy to compare available funding to the amount of maintenance work that actually needs to be completed. Through use of this visualization, it was discovered that an inappropriate amount of money was being allocated for maintenance. The visualization also allowed this dis- covery to be effectively communicated to the state legislature strictly in terms of data, allowing them to use that data in reconsidering more efficient allocation of funds. Another story is of one Gray Notebook visualization, seen in Table 5, created to communicate mobility performance that carries the headline “Incident Response teams provides an estimated $20.5 million in economic benefit.” The visualization uses a chart to represent change over time Figure 42. Line chart example (Source: WSDOT). Table 5. Table example from WSDOT.

58 Visualization of Highway Performance Measures in total number of incidents, percent blocking, average incident clearing time, cost of incident induced delay, economic benefits of Incident Response, and percentage of change from the previous year of 2019. This visualization was able to present a very clear data story, in which the data showed the economic benefit of investment in the Incident Response Team program in terms of dollars and cents. Again, this visualization was able to communicate this performance story simply and effectively to the state legislature, which, in turn, decided to invest additional resources into the Incident Response Team program. Complex Chart Types If an analyst is looking at big data, and trying to see what the data are saying, then a very com- plex visualization can be the most effective tool to find the “needle in the haystack.” Figure 43, from WSDOT’s COVID interactive dashboard, shows how even a more complex graphic with lots of data continues to follow the principles of simplicity. The audience has control over the extent of the data shown by geography, time scale, and category of crash trend. Video As an ultimate storytelling device, including a video in a webpage or dashboard, as seen in Fig- ure 44, can be a powerful tool for communicating performance stories. Responding state DOTs reported that typically, the shorter the video, the more likes and views it receives. A 2-minute video that is shared by many people has far more impact than a 20-minute video that is only shared by a few. WSDOT Multimodal Mobility Dashboard Built on the success of the Gray Notebook, this dashboard is designed to provide specific infor- mation about eight measures in five different regions. Starting from an executive summary, the individual data sheets are shown for each measure. The information can be viewed at a cor- ridor level, all the way down to the data at a specific mile point. The deeper down one goes on the pyramid, the more sophisticated the visualizations become. As shown in Figure 45, the dashboard starts with showing a summary of information (like seeing the whole forest) and then allows the user to drill down to the deepest level required to find information to answer their question (like seeing the leaf). This process allows WSDOT an opportunity to educate the audience as they explore the data freely. Figure 46 shows how the content from the dashboard fits into this structure. The Executive Summary shown in Figure 47 visualizes the “forest”, summarizing information about the eight modes reported in the dashboard. The total values and trending directions are simply presented in a graphical table. By providing increasing levels of detail as the user drills down into deeper levels of the dashboard, the user is given more context. The heat map shown in Figure 48 represents the kind of “leaf” data visualized at the bottom of the pyramid. This more complex visualization from the Wash- ington Multimodal Mobility Dashboard shows congestion in morning and afternoon hours, by location shown on an interactive color-coded map. A user can highlight a specific road segment and view specific information in context. WSDOT visualizations use simple language. WSDOT verifies the data are accurate and that information is visualized through charts, graphs, and infographics to inform their audience. This dashboard is a valuable companion to the Gray Notebook. The performance visualizations facilitate communication that helps to break the silos between various divisions or agencies and in so doing demonstrates the potential power of visualizing performance stories, whether they be positive or negative. Regardless of the nature of the story, the data are reported accurately and transparently.

Figure 43. Example of more complex visualization from dashboard (Source: WSDOT).

Figure 44. Video example (Source: WSDOT).

Case Examples 61   Figure 45. Generic dashboard structure. EXECUTIVE SUMMARY INDIVIDUAL MEASURE SUMMARY REGION EXECUTIVE SUMMARY DETAILED ANALYSIS 8 M eas u res + 5 regions V iew dat a s heet s for each meas u re D et ail ed info for corridors Heat map s and s p ot dat a Figure 46. Multimodal mobility dashboard (Source: WSDOT).

62 Visualization of Highway Performance Measures Figure 47. Executive summary begins dashboard (Source: WSDOT n.d.).

Case Examples 63   Figure 48. Heat map detail from dashboard (Source: WSDOT).

64 Visualization of Highway Performance Measures Virginia DOT: Dashboard Makes an Impact Virginia DOT has had success in using Power BI and SharePoint software to create visual- ization tools to communicate performance measures in transportation construction projects, safety, and road conditions to the general public. Internally, they have used the same tools to create a myriad of dashboards, as shown in Figure 49, capable of delivering visualizations of increasingly specified data tailored to the needs of individual divisions and agencies. The signifi- cance of VDOT’s success is that it has shown that the more successful visualization tools that are created, the more individuals, businesses, and government agencies are finding them useful for decision making and building greater overall trust in the system through data transparency. One VDOT practitioner noted that there are now so many visualization tools available within the VDOT system that one of their next challenges is establishing governance for creating reports using these tools to prevent the creation of seemingly conflicting reports due to the lack of uni- form standards. VDOT began its efforts at visualizing performance measures in 2003. In 2002, only 20% of projects were completed on time. The decision was made to publish a dashboard that identified current progress of planning and construction projects. With a belief that “what gets measured, gets done,” VDOT created a performance dashboard that included performance, safety, condition, finance, management, projects, and citizen survey results. Project on-time delivery improved dramatically after the dashboard was published. The opening screen, seen in Figure 50, shows seven performance measures composed of four highway measures and three VDOT performance measures, as seen in Figure 51. Each of those measures provides more detailed data visualizations, many of which can be shown for the specific area or category desired. The appearance of this website is intentionally designed to visually engage at the level of the general public. It is designed to look like a vehicle dashboard, with needles indicating status of a particular measure, reinforced by green and red color-coded sections of each gauge that communicate performance “at-a-glance.” This simple design presents two sets of gauges, one for highway performance measures and one for VDOT management performance measures. The highway performance categories represented include traffic performance, safety in terms of year-to-date highway deaths, road condition, and finance. As shown in Figure 52, the VDOT management measures represented include management, citizen survey results, and project development and delivery. All of the information previously available via their previous SMART SCALE dashboard is now folded into the larger VDOT Dashboard, allowing for the phasing out of the SMART SCALE dashboard in the immediate future. The user can click any of the gauges to be taken to more detailed information with additional visualization tools communicating more specific data. Each category provides its own specified gauges for relevant measures, as well as pie charts, graphs, and tables that can be individualized through the use of filters. For example, clicking the Safety tab will take the user to a site that presents gauges visualizing crashes injuries, year-to-date deaths, and Work Zone crashes. The site also contains a chart showing data for type of crash, such as deer collision or sideswipe, and figures for the most recent 12 months and 3-year average. These data can be filtered by easily checking boxes on a drop-down menu, selecting for county, district, or statewide. Those inter- viewed expressed that because of the transparency that the dashboard provides, VDOT’s ability to deliver has not been questioned in almost 20 years. That effort was expanded about 4 or 5 years ago with the SMART SCALE dashboard project, illustrated in Figure 53. At the time, effective data visualization software such as Power BI and SharePoint was not available. Therefore, support of time and resources for the approximately $600,000 project was crucial. A significant part of the cost of the project was paying for the

Figure 49. Virginia DOT dashboard opening screen and projects detail (Source: VDOT).

Figure 50. Opening screen Virginia dashboard (updated in 2021) (Source: VDOT 2022).

Case Examples 67   7 PERFORMANCE MEASURES 3 PERFORMANCE MEASURES 4 HIGHWAY MEASURES MORE DETAILED 4 Highway and 3 Performance Management | Citizen Survey | Projects Performance | Safety Condition | Finance Make selection Figure 51. Structure of Virginia dashboard performance reporting system (Source: VDOT). Java coding expertise to create the dashboard. e SMART SCALE dashboard was created speci- cally for the purpose of tracking construction projects that would be funded using VDOT’s new SMART SCALE program. is program initiated a process of prioritizing funding of certain new construction projects. e dashboard was designed to track and visualize the per- formance of each project in terms of progress toward completion milestones and remaining within budget. Projects are then given a color code of green, yellow, or red to indicate their performance status. ese data are visualized in several ways. ere are circular gauges that visualize the percentage of overall projects that are on time and projects that are on budget. ere is also an interactive map that visualizes project locations, which can be clicked to lead to more detailed information about each project and its progress toward budget and completion. At the same time that VDOT introduced the SMART SCALE program and dashboard, VDOT also implemented new business rules to coincide with the scoring and performance metrics. VDOT’s approach to audience emphasizes what VDOT believes is a major factor in its success. Regardless of the intended use of the visualization tool, the targeted audience always remains the general public. e reasoning given for this approach can be said to be one of the guiding prin- ciples of VDOT’s perspective on visualization: the primary goal is simplicity. If a visualization is designed to communicate data in a manner simple enough to be understood by the general public, then it is an eective informational tool for anyone who may need to use it. is does not mean that they do not create more complex and specic visualizations for internal audiences; however, they adhere to the overall goal of simplicity as the gold standard. e SMART SCALE dashboard proved to be useful and productive in raising the awareness of construction project managers of their status regarding key performance measures, improving management of both time and budget on some projects, and building trust within the system and with the public through providing transparency with real-time data. e success of the SMART SCALE dashboard was built on the design and implementation of a much larger scale visualization project, the VDOT Performance Reporting System for Projects and Programs. e success of this visualization project cannot be overstated. VDOT’s credit- ability was rebuilt around their ability to develop and deliver projects on time and on budget.

Figure 52. Detail on on-time, on-budget performance from the Virginia dashboard (Source: VDOT).

Case Examples 69   In 2002, only 20% of projects were completed on time. In 2003, VDOT created a dashboard that reported on the on-time, on-budget progress of projects. VDOT has consistently met its performance target of 77% for most of the last 10 years. VDOT representatives expressed that performance has been critical to instilling confidence in the DOT and the success of several funding initiatives. Again, those interviewed from VDOT emphasized their perception that it is because of the transparency that the dashboard provides that VDOT’s ability to deliver has not been questioned in almost 20 years. VDOT has been able to use data tracked through these projects to identify inefficiencies and design targeted solutions. For example, VDOT was able to track the performance of VDOT- administered construction projects compared to projects they had permitted to be locally admin- istered. They discovered that locally administered projects performed lower in on-time and on-budget metrics than VDOT-administered projects. In response, they implemented an internal visualization tool of dashboards designed to target locally administered project performance Figure 53. SMART SCALE dashboard (Source: VDOT).

70 Visualization of Highway Performance Measures measures. Most significantly, these local dashboards visualize the current status of active projects, showing at what point in the development and implementation stages of a project it runs into trouble and why. Visualization of this performance measure has allowed VDOT to pursue one of its next visualization projects, which is attempting to identify which indicators may be predictive of project failure, but more importantly, which indicators are predictive of troubled projects that may be able to be successfully recovered, and how. VDOT is also able to use the performance data tracked and visualized on local dashboards to determine which localities are most likely to complete an administered project successfully and assign responsibility for future construction projects in each locality accordingly, based on overall performance. VDOT attributes their success not only to their dedication to simplicity, but also to their determination to assure that they hire the appropriately qualified people. VDOT representatives responded that it is essential to have the resources to hire people who understand the transpor- tation industry, understand the data, and understand the technological tools being used. It was found that outsourcing technological work to those who do not understand the transportation industry or the data was not productive, nor was attempting to train data experts on the work- ings of the transportation industry. Understandably, identifying and retaining the people with the appropriate skills for the work is a challenge; however, with the apparent rising demand for transportation data visualization tools and number of projects underway within VDOT alone, there is reason to anticipate that the number of qualified and experienced personnel in this field may be on the rise. VDOT’s experience is strong evidence of how the creation of successful visualization tools for performance measures can translate to the identification of real-world problems and lead to further innovation of visualization to assist in solving them. Florida DOT: Defining Multimodal Mobility Performance Measures Another DOT that has been a pioneer in the visualization of transportation data is Florida. FDOT has been exploring ways to better communicate performance measures since the late 1990’s, when they began working with about a dozen roadway measures. However, the most sig- nificant innovation of FDOT’s efforts has been their vision of visualizing data both horizontally (showcasing multiple modes) and vertically (ranging from local to state level). They have built partnerships within the transportation system to make decisions using the visualization of multi- modal data. More importantly, these visualization tools are currently being used to increase safety and efficiency in all modes of mobility. FDOT’s extensive efforts have produced the digi- tally accessible FDOT Source Book, a sample of which is shown in Figure 54. This website is an excellent example of what is possible with data visualization. It provides visualization of multimodal data in easily navigable categories such as mobility, factors affecting mobility, infra- structure, safety, and accountability. Users can select the data they need and access it in an easy- to-understand visualized format. This allows previously inaccessible data collected by various agencies and localities to now be used to increase efficiency, safety, and resource allocation horizontally across multiple modes as well as vertically, from the local to the state level. FDOT has a clear understanding of which audiences they are targeting with their visualiza- tion tools and how to prioritize them. Planners are their first priority, with significant consid- eration of the data needs of the larger transportation community, and then the general public. Partner ships with districts and other localities within the state also allow for consideration of specific audiences on local levels, such as MPOs and other transportation partners. A district- level Senior Policy Planning Analyst in the state commented that the question is always “How do you broaden the constituency of those who can benefit from the availability of this informa- tion?” The primary audience must be clear to create the most effective visualization tools, while also keeping in mind ways that the tools could and may be used by a wider audience.

Figure 54. Example of multimodal mobility performance measures from The FDOT Source Book (Source: FDOT).

72 Visualization of Highway Performance Measures It is important to note that responding state DOTs emphasize the significance of both demand from organizations in the transportation community, as well as considerable support from executive and government agencies to their success. The current FDOT Emerging Transporta- tion Coordinator in Forecasting and Trends states that the initial impetus for the Source Book project came from requests from customers for the availability of this type of data in an easily accessible and usable form. It was the constructive communication of this demand to executive teams that helped to bolster support in terms of providing the appropriate allocation of time and resources necessary to complete a project of this size effectively. Due to visualization tools relying on both complete and accurate data collection and organization, the Source Book was not a project that could be done in pieces. It was essential to have the support and resources available to ensure that the project could be effectively sustained to completion and regularly maintained. Due to the sustained support FDOT and other local agencies received, they have been able to form productive working partnerships to assist with both data collection and organization. Districts and localities have been proving to be a fertile testing ground for visualization projects that can then be translated to implementation at higher levels. They have also been able to form similar partnerships between the agencies governing the various modes of mobility within the state. This level of cooperation and communication has now created the potential for a fully inte- grated transportation system in which data can be shared across modalities, allowing for multi- modal solutions to a broad spectrum of challenges to mobility throughout the state. Finally, the success of the Source Book has created an increasing demand from an ever-wider audience of users for more and more data visualization tools, indicating a growing need for continued innovation in the field. Utah DOT: Telling the Utah Performance Story Similar to other DOTs that were interviewed, Utah DOT (UDOT) was originally reporting its performance data and visualizations in an annually published Strategic Direction Report. The impetus for innovation into the use of visualization to report performance data came from the top down, in the sense that leadership agreed that it needed to be done and were willing and able to provide the necessary support for it to be done. However, how to do so began to proliferate from the bottom up, as more and more agencies and divisions began hiring their own GIS personnel to do it. This ultimately resulted in a relatively happy marriage in the middle, where UDOT has been able to establish certain criteria of governance over pro- cess and procedure, as well as a method for centralizing data collection in the areas they are currently monitoring on a statewide basis. Regional information is also available through UDOT’s reporting tools, but it is important to note that there are many regional and local visualization tools that are being created and used solely by those localities internally that are not currently linked into the visualization tools that are accessible publicly or system wide. This partnership between regional and central personnel as well as a strong partnership between the UDOT Division of Transportation and Performance Management and the relatively newly created UDOT Division of Data, Technology, and Analytics has allowed for this formerly static docu- ment to be converted to a dynamic “living document” that is capable of providing visualiza- tions of real-time performance metrics, allowing decision makers to make more informed and productive decisions. The Strategic Direction website, shown in Figure 55, serves as UDOT’s “face” to decision makers and other transportation system stakeholders. When questions are asked, not only regarding performance measure data, but also about UDOT as a whole, they can easily be answered by a referral to the Strategic Direction website. The site reflects this perspective in its design.

Figure 55. Utah dashboard (Source: UDOT).

74 Visualization of Highway Performance Measures Figure 56. UDOT dashboard splash screen (Source: UDOT). From a simple splash screen that says “KEEPING UTAH MOVING,” as seen in Figure 56, the page scrolls to a high-level summary of strategic goals. These are three strategically selected measures of safety, mobility, and infrastructure, as well as funding and projects. As indicated in Figure 57, each of those categories leads to additional detail, more dashboards, and finally to a data-rich mobility dashboard that can visualize mobility data from statewide down to a single bottleneck (from the forest to the leaf). UDOT has clearly defined their audiences and how the DOT prioritizes audiences. The Gov- ernor’s Office and the legislature are their designated primary audience, while they have defined the general public (as investors in the system) and internal users as ancillary. Therefore, it was established early on that all visualization tools must be able to present relevant data in a form that clearly and simply communicated the most significant point, or performance story, in the most effectively direct manner possible. Figure 57. Structure of Utah strategic direction dashboard (Source: UDOT).

Case Examples 75   The public-facing homepage for the site, shown in Figure 58, emphasizes the mission of UDOT in a logical, general-to-specific way. The first links available provide information about their over- all mission, vision, values, and strategic goals, often presented in an infographic format. There is also a link that takes the user to a sleekly designed document titled “Getting to Know UDOT,” shown in Figure 59. Using text, photographs, and graphics, as a traditionally designed document, it walks through all of the different functions UDOT serves throughout the state and is updated periodically as the system evolves. Figure 60 shows that the left side of the homepage is an interactive menu presenting the general categories of information contained throughout the site, represented in both text and graphic forms. Figure 58. UDOT dashboard road map (Source: UDOT). Figure 59. UDOT overview brochure (Source: UDOT).

Figure 60. UDOT dashboard high-level overview (Source: UDOT).

Case Examples 77   For example, the section on Funding is labeled with both the word and a clickable graphic of stacks of coins. The general categories available are Funding, Projects, and Tactical Measures. Each of these categories serves as a gateway to increasingly granular data, all of which are visual- ized and, in many cases, also downloadable for practical application. Figure 61 shows an example, with visualizations and data related to revenue. One of the most significant characteristics of UDOT’s experience is the number of examples the DOT can provide of ways visualizing data has had an impact on decision-making processes legislatively and at the level of individual human behavior. The Mobility Dashboard, shown in Figure 62, provides detailed information that provides multiple tabs and user options for many different elements to provide a detailed visualization of mobility from the statewide forest view to the leaf view of a single bottleneck. One richly complex example is UDOT’s Zero Fatality section under its category of Strategic Goals, as shown in Figure 63. The phrase “Zero Fatalities” can be used as an umbrella that rep- resents an overall safety goal that nearly all people can agree on; everyone understands the goal of no fatalities. The use of safety data in a wide variety of visualized formats has helped develop an approach of using data visualization to inform a wider program of action. It allows for the use of data to help identify specific problems and innovate targeted solution strategies with an increased likelihood of effectiveness. Under Strategic Goals, the various safety measures are represented as bar graphs reflecting how close UDOT is to meeting 100% of all safety targets, and the overall goal is to have all indices rise to 100. A line chart showing change over time shows a significant decrease in the overall index since 2019. This visualization tells the performance story that overall system safety is decreasing, identifying clearly that a problem exists and allowing energy to be directed next to answering the question of why this is happening. Clicking on the Zero Fatalities section of this category takes the user to visualizations of more detailed data. A combined bar graph and line chart representing fatality rate per 100 miles VMT between 2015 and 2021 shows that VMT decreased during the pandemic, but fatalities increased. They are now seeing VMT increase while the higher fatality rate is remaining the same. The next question then becomes “What data metrics are available to help further illumi- nate what is actually happening on the ground?” While the tools to overlay the necessary metrics to potentially visualize a more targeted answer do not yet exist among respondents, UDOT has already developed an impressive array of tools that can help to narrow down the areas in which time, energy, and other resources have an increased likelihood for effectiveness. While one must still spend a good deal of time exploring various sections of the website to find all the data necessary to attempt to make these determinations, it is an accomplishment to have streamlined the data analysis and communication function to simply navigating a user-friendly, interactive website. UDOT specifically identifies the streamlining of communication as a notice- able benefit of having this site. It allows person-to-person time to be more focused on explana- tion, analysis, and solution strategy planning as opposed to how to locate data and build necessary reports. There is a section on strategies that provides data visualizations in the categories of Infra- structure Improvement, Partnerships with Public Safety, Employee Safety, and Public Outreach and Education. Line charts are available representing six major safety performance measures: Traffic Fatalities, Internal Fatalities, Suspected Serious Traffic Injuries, Internal Injury Rate, Traffic Crashes, and Equipment Damage Rate. Monitoring of these visualizations has allowed for more compelling data-based discussions for including improved safety features in projects. However, even more granular data are available on the level of both regions and MPOs. The Strategic Direction website has a section dedicated specifically to reporting on Federal Perfor- mance Measures, and in this section, there are tools provided to help MPOs and regions track their relative progress toward federal targets and therefore add an additional level of focus to the flow of appropriate resources to the appropriate places on the localized level.

Figure 61. UDOT dashboard revenue summary (Source: UDOT).

Figure 62. UDOT mobility dashboard for Region 2 (Source: UDOT).

Figure 63. Zero fatalities performance measure (Source: UDOT).

Case Examples 81   What can be determined by UDOT’s available data is that the roads within UDOT’s jurisdic- tion did not suffer any comparable drop in condition or other safety measure that would explain the continued higher fatality rate they are currently experiencing. This identification serves to help guide the direction for further innovation in how to create more complex tools that allow for the telling of even more detailed data stories. It also helps to allow for a more informed level of theorizing. For example, if there does not seem to be any corresponding trend in the available data, it becomes more reasonable to invest time and other resources into looking for answers in areas that are not currently measured, such as human behavior. It is possible that the cause could lie in human behaviors such as increased numbers of people speeding or engaging in distracted driving behaviors like texting. If a problem, such as sustained increased fatality rate despite rising VMT, could be reasonably deduced to possibly being an issue of human behavior, UDOT already has experience in tackling the complex question of how to attempt to impact dangerous human behavior. When it was identified that drivers were 23 times more likely to die in a crash while texting, UDOT turned to a creative but controversial method of visually communicating that data to the public. In 2018, they invested in a Super Bowl advertisement that dramatized a family in a vehicle, with parents sitting in the front seat and their young daughter in the backseat, as the vehicle is impacted in slow motion. The view is of the family inside the car. As glass begins to shatter and the parents lurch sideways with the force of the impact, focus zooms in on the child in the backseat who sits up straight, in the midst of it all, and speaks directly to the camera as she assertively says: “This is your fault.” The data about the danger of texting and driving are then presented, leaving a dramatic impact on the viewer. UDOT has partially credited the discussion and awareness of crash safety generated by this advertisement to the passage of a primary seatbelt law in recent years and reports that there are some positive indicators that compliance may be rising, particu- larly in rural areas. While the Zero Fatalities data are not yet capable of providing an entirely data-based answer to why the fatality rate is currently still so high throughout the system, it has provided signifi- cant insight into where best to direct valuable resources toward more specifically identifying the problem, as well as narrower direction and potential strategy options for their efforts to address the problem in the meantime. More importantly, having the data visualization tools easily avail- able allows for the spread of communication and an awareness of the issue throughout the stakeholder community. The area in which UDOT identifies the largest number of practical applications is in that of budgeting. Many of their tools for visualization performance in this area are internally facing. This decision is attributed to the complexity of the data involved. There is a set of available tools for Maintenance and Facility Management that allows the user to access a wide variety of budgetary information. For example, for their overall Maintenance budget they have both dials and bar charts that represent Current Fiscal Year Budget, Projected Expenditures, and percentage of budget left. Detailed tables are also provided for each project. Just having these simple visu- alizations available makes it much easier to keep everyone on the same page in terms of under- standing what money is being spent, and how close they are to maintaining their budgets. Most significantly, these visualizations provide the basis for much more meaningful and productive conversations between stakeholders than they have found to be possible using spreadsheets. UDOT identifies others of its visualization tools as also being particularly useful for budgeting purposes, such as in internal project management. For example, one tool visualizes expenditures against project value. It measures the number of projects that are at less than 95% of project value, the number within 5% of project value, and the number of projects over project value. This visual currently communicates that they currently have zero projects over project value; how- ever, any change in this indicator can allow for an expedient examination into possible causes,

82 Visualization of Highway Performance Measures and appropriate remedial strategies can more quickly be identified and implemented. An even simpler and practically useful visualization that they have found helpful in budget forecasting is a set of tools designed as dials that communicate the total amount of the annual UDOT snow budget for the year, the amount used, and the amount remaining. UDOT policy is to divide any remaining funds allocated for responding to snow events among the regions to be used for main- tenance needs. This tool allows regions to monitor the use of the snow budget funds throughout the year and can therefore more accurately anticipate whether they may have any additional funding from this source, and if so, how much it is likely to be. UDOT also provides productive insight to the evolutionary nature of collecting, governing, and designing visualization tools that function as intended. They found it is important that for data and subsequent visualizations to be useful, the appropriate question should be asked in the appropriate manner. For example, in their section on preserving infrastructure, they created a visualization tool that was intended to measure the overall status of the roadways toward meet- ing performance targets. The performance story that the data were telling over the course of several years was that the roadways were meeting 100% of their targets, implying that no addi- tional money was needed and that the roadways were all in good condition. This did not accu- rately reflect the reality of the situation. The problem was then identified that the metric that was being used to collect data from throughout the system was formulated as a simple yes or no metric: “Did you meet your target?” As a result, in 2019 the DOT changed the formula used to compute the metric to include pavement conditions more accurately, therefore making the data significantly more useful in determining and predicting when and how much more funding will be required to reach specified targets. UDOT has created a dynamic site that allows them to represent their goals, work, and needs accurately and transparently in an easy-to-navigate and easy-to-understand format that they have found makes it easier to identify and solve problems as well as to communicate with and influence the decision-making community to make the system safer and more efficient overall. Arkansas DOT: Following the Leader to Lead by Example It is appropriate to conclude the case examples with Arkansas DOT (ARDOT), because their experience illustrates the benefits that can now be gained by learning from the successes and pitfalls of others. Initial support for increasing the visualization of performance data came from the ARDOT leadership down. However, as has been seen in other DOTs, the how to do so has come from innovating thinkers, particularly in the Planning and Maintenance Divisions across ARDOT. ARDOT respondents were forthcoming about their reliance on building on the successes of what other DOTs, such as UDOT, were beginning to produce. It is significant to emphasize that one of the early steps ARDOT took in their development process was to form a data governance team, an issue that has been seen to be a challenge in the previous case examples. Around 2014, they really began to turn their focus toward using digital and online tools and platforms. Continued support from leadership has manifested in many ways, includ- ing open channels of communication, in which leadership is interested in seeing the products that are being created, as well as offering advice. Another challenge identified by other DOTs is the issue of sustained support and maintenance in order for these tools to remain useful. This is an issue that ARDOT respondents spoke frequently on, and they reported that it is an ongoing priority of the supporting leadership of the department. ARDOT’s experience reflects a department that is appropriately concerned with proceeding in the most effective and efficient manner and is eager to make use of all available lessons learned to do so. ARDOT has already created a number of visualization tools available on public websites and internally and is working toward the goal of creating its own centralized performance manage- ment dashboard, as illustrated in Figure 64, similar to those presented in the other case examples.

Figure 64. Arkansas DOT bridges dashboard (Source: ARDOT 2022).

84 Visualization of Highway Performance Measures ARDOT respondents currently working on performance visualization within ARDOT all com- mented extensively on the deliberateness with which they are thinking about what types of data to visualize, in what specific manner, differences between internal and external usage and audiences, and the need for governance of data collection, storage, and usage. A learning process is begin- ning to take place as different DOTs are publishing more and more of their work in this area, and these are the important considerations that those at ARDOT are talking most about as they look toward creating their own centralized performance management dashboard. ARDOT identifies that their intended audience includes two major types. The primary audi- ence is the decision makers, in order to inform them about the current performance of the department, and to communicate to them any gap between current performance and targets. Secondly, there is the public, in order to improve the transparency of the DOT and to be account- able for how they are spending taxpayers’ money. However, they also acknowledge there is a significant difference in designing visualization tools for the public versus for internal use, and a large determining factor in that design is relative complexity. Their current internal tools con- tain data and visualizations that require a certain level of familiarity with the subject matter to be understood, while their public tools reflect an increased attention to clarity and simplicity. Two general areas of performance data visualization that are emphasized by ARDOT are traffic/ safety and bridge maintenance. Additionally, a large amount of information is available through ARDOT’s website that feature Virtual Public Involvement Meetings. These were an ideal way to reach Arkansans during the COVID-19 pandemic. The site contains public surveys for collecting data about public opinion, videos, interactive maps, photographs, and other information about upcoming projects. An ARDOT practitioner explained their Crash Analytics Dashboard, shown in Figure 65, in detail. This publicly facing dashboard allows a very large amount of data to be filtered and visual- ized in a variety of ways, from the general to the increasingly specific. The user can easily select from a number of performance categories on tabs at the top of the page, including Roadway Departure Crashes, Impaired Driver Crashes, Speed Related Crashes, Young Driver Related Crashes, Intersection Related Crashes, Work Zone Related Crashes and others. The page on Fatal and Suspected Serious Injury Crashes contains an interactive map showing crash locations, while next to it, in the upper right-hand corner, numbers are given for total crashes, fatal crashes, and suspected serious injury crashes. Bar graphs are also shown representing Fatal Crashes by Emphasis and Suspected Serious Injury Crashes by Emphasis. The goal for the Crash Analytics Dashboard project is to ultimately show the location and severity of every crash throughout Arkansas; however, one of the challenges that ARDOT says they are facing toward this goal is that not all agencies have adopted the eCrash system that is being used to collect and govern this data. They were, however, able to successfully navigate the challenge of creating the necessary data partnership and memorandum of understanding with the Arkansas State Police that allows them access to the necessary databases to create and main- tain this tool, and to do so while protecting sensitive personal and legal information that should not be shared publicly. ARDOT also expressed that they are finding that as increasing numbers of agencies are using the eCrash system and making use of the performance data available on the site, it has been encouraging other agencies and stakeholders to also begin using it as well, increasing the accuracy and effectiveness of the data visualized on the site. Another way in which various types of traffic performance data are being visualized by ARDOT is through the traveler information website IDrive Arkansas, which was first launched in 2013 (Arkansas Department of Transportation n.d.). Originally, the site just contained infor- mation about the locations of current construction projects. However, today the site provides an interactive map with an array of filters to select for which types of information the user

Figure 65. ACT (Arkansas Crash Analysis Tool) (Source: ARDOT).

86 Visualization of Highway Performance Measures would like to visualize. Construction project locations can still be represented as well as weight- restricted routes, rest areas, and feeds from live traffic cameras throughout the ARDOT system. IDrive also provides a link for user feedback, through which ARDOT can identify priorities of the members of the public, such as where to add new traffic cameras or information on the adopt-a-highway program. An application was also used through this site after a particularly damaging winter storm that allowed for the public to identify the locations of potholes, helping the department better target road maintenance efforts. The other area that ARDOT reports significant progress in is through their Bridge Data Dash- board, which currently has both internally and externally facing tools. The publicly facing site contains information about the total number of bridges, an interactive map of bridge locations that allows the user to click on the location and get more detailed information about it as well as a photograph, and the number of bridges that in are in good, fair, and poor condition. ARDOT comments that it is through sites like this that they feel they are able to communicate to the public that ARDOT is aware of their concerns, such as aging infrastructure, and that ARDOT is actively working to address them. These tools are currently reported to be having the most impact internally. The internally facing tools contain a markedly more complex level of data and available filters. One tool in particular is emphasized that visualizes program bridge data with inspection data. An ARDOT Heavy Bridge Maintenance Engineer estimates that this visualization has made it exponentially faster for the department to identify where and when specific bridges need to be replaced and better manage necessary resources in the appro- priate manner. ARDOT is now currently working toward the development of its own centralized performance management dashboard, with an emphasis placed by them on centralization. While they are actively experimenting with what types of data to visualize outside federally established perfor- mance measures and how best to visualize that data, the challenge that is expressed emphati- cally is the need to develop universal governance policies surrounding the collection of data and how data are used to create visualizations throughout the ARDOT system. They are excited to report that they know their applications are being used, particularly their daily traffic tools and their bridge information tools, because they receive considerable feedback whenever there happens to be a glitch in the system. However, this also communicates that these tools are increasingly being relied on by stakeholders in the transportation system, which emphasizes to them the need to proceed carefully and ensure that the tools that are being created are as accurate, transparent, and consistently maintained as possible from the beginning. Observations about the Target Audiences from the Interviews During the interviews, the DOTs responded about who their primary and secondary audience is for the performance measures that they visualize. Table 6 summarizes the different perspec- tives for each state interviewed as described in the text below. The explanations provide insight into these state DOT approaches to visualizing performance measures. WSDOT’s intent and purpose from the earliest years is to speak to those who could most use this information to inform decisions, particularly in terms of where to best allocate efforts and resources, and which efforts and resources would be most effective. This audience is defined as the general public, elected officials, and other stakeholders in the transportation community. VDOT’s approach to audience emphasizes a factor they have identified as key to their success. Regardless of the intended use of the visualization tool, the targeted audience always remains

Case Examples 87   the general public. The reasoning given for this approach can be said to be one of the guiding principles of VDOT’s perspective on visualization; the primary goal is simplicity. If a visual- ization is designed to communicate data in a manner simple enough to be understood by the general public, then it is an effective informational tool for anyone who may need to use it. This does not mean that VDOT does not create more complex and specified visualizations for internal audiences, rather that the DOT adheres to the overall goal of simplicity as the gold standard. Internal DOT Internal Leadership Elected Public Official External Audience WS General Public VA General Public FL Planners MPOs UT Governor/Legislature AR Decision Makers legend Primary Audience Primary Secondary Audience Secondary Additional Audience Table 6. Explanation of primary and secondary audience from state interviews.