The University of Maryland’s Center for Advanced Transportation Technology Laboratory has created a number of  Probe Data Analytics  (PDA) apps that give users tremendous  capabilities to evaluate system performance in support of planning, operations, research and other activities. PDA uses 3rd party probe data fused with other agency transportation data that provides a true “big data” analytics platform. These data are then leveraged by nine different apps to generate actionable data summaries, maps, graphs, charts and unique visualizations for detailed analyses and reporting. Clear, concise and easy-to-understand results are also perfect for publishing press releases for public and media consumption.

What are the worst bottlenecks?

What are the worst bottlenecks?

Where is the most congestion?

Where is the most congestion?

Some of the benefits of using PDA include:

  • Enhanced real-time operations decision-making
  • Better work zone monitoring
  • Improved travel time and reliability analysis
  • More robust After Action Reviews
  • Comprehensive problem identification
  • Cost-effective project/program prioritization
  • Improved Before & After Studies
    How did that storm impact travel, compared to other normal days?

    How did that storm impact travel, compared to other normal days?

  • Extensive system performance reporting capabilities
  • Demonstrate the economic impacts of user delay
  • Fully compliant with MAP-21 reporting requirements

To help facilitate PDA use and improve features and functionality, a User Group was formed in 2012. The group is comprised of “power users” from Coalition membership with demonstrated skills in using PDA in effective and creative ways. The group meets quarterly to share unique PDA uses, exchange ideas and work collaboratively with the developers to ensure PDA apps meet current and future needs.

For additional information on PDA, please contact
Michael L. Pack at PackML@umd.edu.

Quick Links: Quick Start Sheet, PDA Tutorials, PDA Log-in, Probe Data Analytics Forum

Meeting Info (slides/ minutes) from previous meetings

O-D Data Suite Focus Group

The Probe Data Analytics Forum is a “User Hub” developed for agencies using the CATT Lab’s PDA apps. Use this site for posting and sharing information and attachments – like Agency Press Releases, Agency Use Cases and Tutorials – that helps make PDA use easier and more effective. You can hold technical discussions, participate in User Groups and in general coordinate and collaborate in a “one-stop shop” fashion on any and all things PDA!

Using the Probe Data Analytics Forum will make your PDA use and the products you develop smarter and better; ultimately saving you and your organization time, money and effort.

It’s easy to become a member – just click here, then click on Register in the upper right hand corner of the Probe Data Analytics Forum home page, and you’ll be up and running in no time!

forum-screenshot

Data Agnostic: Probe Data Analytics works with all agency data plus 3rd party data from ATMS vendors, ITS device manufacturers, and data providers like HERE, INRIX, and TomTom data.  You can even change providers and keep your historic data.  Similarly, our system allows you to explore data and perform analysis outside of your agency’s borders.  There is great value to seeing across our jurisdictional boundaries.

Engaged in and Committed to MAP-21:  CATT Lab developers, management, and user group members are tuned-in to MAP-21 rulemaking efforts, and have provided significant guidance to FHWA.  The team has been preparing and doing work in the background to ensure PDA apps meet your MAP-21 target setting goals as soon as the final rules are published

Open and Transparent:  Want to know how we calculate BIT, TTI, bottlenecks, and UDC?  All of our formulas are open to the public, and our users (including agency transportation engineers, academics, and FHWA partners) helped us define, review, and approve them.

Unparalleled Team Support:  PDA apps are developed and supported by a dedicated interdisciplinary team of graphic artists, UX specialists, transportation faculty, IT staff, professional engineers, software developers, and agency thought leaders—making these tools a powerful national asset.

Agency driven enhancements and feature development:  Agency analysts, statisticians, managers, and operations specialists drive our software development goals.  You have a voice in what we build into the system next.

Resource and costs sharing:  When one agency invests in a new feature or hardware for PDA apps, all other agencies benefit.  The Suite is truly a shared-resource, pooled fund project.  Working together, we can ultimately accomplish more by leveraging each other’s accomplishments and knowledge to ensure a better, constantly evolving system.

A high-level tutorial video can be found here. However, below are a series of screenshots on some of the more prevalent features.vpp2

 

Time and Location Selections

Figure 1: A screenshot of just one of the four tabs available to help users select the exact road segment(s) or geography needed in any analysis.

Probe Data Analytics has a series of extremely robust filtering capabilities that let you choose the exact location for which you want to perform your analysis. Users can select any level of geography for analysis – from a single segment in a single direction of travel to every single county of multiple states. Users can also select Regions (state and counties), by zip codes and for specific road classes. Complex road selections can be saved for quick and easy retrieval.

Similarly, you can select a single day, multiple days, months, years, etc. for which you want to analyze. You can even select only certain days of the week throughout a date range, or even only certain hours of certain days (Example: morning rush hours on weekdays only).vpp4

Region Explorer

Region explorer shows the relationships between bottlenecks and traffic events and their impacts on traffic conditions both in real-time or at previous points in the past. It is primarily map-based, providing a geospatial context for the data being presented. A ranked list of real-time or point-in-time bottlenecks is shown beside the map, and each element (bottleneck, travel time reading, and/or accident) can be clicked on to bring up additional details, stats, etc. Additionally, historic weather radar data can be viewed.vpp5v2

Massive Raw Data Downloadervpp6v2

All of the raw probe speed data (from any 3rd party provider) can be downloaded using the Massive Raw Data Downloader. This gives the user the ability to do offline, detailed analyses or integrate this raw data into their own applications

Congestion Scan

Congestion Scan lets you analyze traffic conditions on a contiguous stretch of road. If you choose to analyze a single day, traffic events and incidents will also be plotted on the road to help determine causality of the congestion. The congestion scan will also allow you to average multiple days’, months’, or years’ worth of traffic onto a single congestion scan. You can also visualize different time periods side-by-side, invaluable for analyzing impacts of work zones, major events or incidents. You can also do multi-road scans, to congestion impacts along a travel corridor, such as a major commuter or recreational route. The data behind each congestion scan (and all of the other tools in the suite) can be easily exported to Excel—pre-formatted for easy analysis and recreation of specific visualization.

Figure 3: Congestion and other measures can be visualized in both directions of travel along a corridor. Accident and event data can be overlaid on the congestion scan to help determine causality.

Figure 3: Congestion and other measures can be visualized in both directions of travel along a corridor. Accident and event data can be overlaid on the congestion scan to help determine causality.

Figure 4: The Congestion Scan tool allows for detailed zooming, cropping, and custom display configurations for each of its graphs.

Figure 4: The Congestion Scan tool allows for detailed zooming, cropping, and custom display configurations for each of its graphs.

Figure 5: Display multiple date ranges on each scan. This is useful for before and after studies of the impacts of construction projects.

Figure 5: Display multiple date ranges on each scan. This is useful for before and after studies of the impacts of construction projects.

Figure 6: Easily export all Congestion Scan data to a preformatted Excel file.

Figure 6: Easily export all Congestion Scan data to a preformatted Excel file.

Trend Map

Trends map produces animations of roadway performance over time, dynamically displaying changes in speeds, congestion and a number of other measures. With Trend Map, you can choose to analyze a single time period, or multiple time periods, displaying them side-by-side to compare and contrast for any scenarios you choose (weekend vs weekday; seasonal variation; before, during and after events; etc.) There are a number of options at your disposal, such as setting color thresholds, map layouts and displaying traffic events. You can save the results as an Excel file, screenshots and animated GIFs or MP4 movies – great for making highly-informative presentations. And you can even create embed code with the click of a button to add animations into your organization’s website.

  • Figure 7: Up to seven maps can be drawn and animated simultaneously with the Trend map tool.

    Figure 7: Up to seven maps can be drawn and animated simultaneously with the Trend map tool.

Figure 8: Trend maps can be shares as Links, Movies, Animated GIFs, and Interactive Embed on agency webpages.

Figure 8: Trend maps can be shares as Links, Movies, Animated GIFs, and Interactive Embed on agency webpages.

Performance Charts

Performance Charts are bar, line, scatter plot, and candlestick charts representing aggregate conditions across stretches of road. The charts can be grouped by time period or by road directionality.  The currently available performance metrics include:

  • Speeds
  • Travel Times
  • Travel time index
  • Planning Time
  • Planning Time index
  • Buffer Time
  • Buffer Time Index
  • Comparative Speed
  • Congestion
  • Historic Averages for the above

    Figure 9: Screenshot of metric selections and display modes in the performance charts.

    Figure 9: Screenshot of metric selections and display modes in the performance charts.

For each of the above metrics, the user can overlay multiple locations and/or time-ranges for easy comparison. 5th/95th and 25th/75th percentile measurements can also be displayed or hidden with the click of a button.

Figure 10: In this screenshot, average speeds along a stretch of I-270 in Maryland are displayed (dark line), while the 25/75th percentiles are shown in dark gray, and the 5th/95th percentile speeds are shown in light gray.

Figure 10: In this screenshot, average speeds along a stretch of I-270 in Maryland are displayed (dark line), while the 25/75th percentiles are shown in dark gray, and the 5th/95th percentile speeds are shown in light gray.

Performance Summaries

Performance Summaries produce several different tables of metrics grouped by day of week, weekdays, and weekends. The results can be compiled for every hour of the day or for specific time ranges, such as AM and PM peak hours. The reports are grouped by road direction.

Figure 11: Performance summary results for both the morning and afternoon rush hours on I-270.

Figure 11: Performance summary results for both the morning and afternoon rush hours on I-270.

Bottleneck Ranking

Bottleneck Ranking allows you to identify, rank and explore bottleneck locations on a roadway.  The results are presented in a table, a map, and a series of other visualizations with statistics on each bottleneck location, such as overall rank, impact, average max length and daily duration and total duration, as well as all events/incidents that occurred within the space of the bottleneck at any time during the time period being analyzed.

vpp16v2

Selecting a bottleneck location from the table will recenter the map on the relevant location, and the adjacent spiral timeline or line chart will be updated to depict when the location was bottlenecked. Checkboxes on the bottom of the map panel allow you to disable the “highlighting” of the selected bottleneck and turn on or off the ranking numbers next to each map icon. The bottleneck times can be viewed in a table which can also be exported to Excel.

The timeline and line charts in the lower right can be used to determine the temporal patters of congestion. For example, some locations may only become congested during the morning rush hours or the evening rush hours. Event (accident and construction) data on top of these graphics helps to determine congestion.

vpp17v2

vpp18v2

Clicking on any element in the resulting table, time spiral graph, or line charts will open up congestion scans for each congestion event and/or draw an incident timeline graphic depicting the event.

User Delay Cost Analytics

User Delay Cost (UDC) Analysis combines probe speed data with volume data provided by the Texas Transportation Institute and/or individual agency volume data to estimate the cost of delay experienced by drivers as a result of congestion. Default values of  commercial and passenger vehicle costs are provided, or you can adjust  based on your own information. You can also adjust the percentages of passenger and commercial vehicles.

Users can define when delay should be calculated based on the following options:

  • Historic Average Speeds: Calculate delay anytime a segment falls below historic average conditions. You can even define how far below historic average you must be before you begin computing delay:
  • Free-flow speeds: You can begin calculating delay anytime the speeds drop below free-flow or some deviation from free-flow (example: 20 mph below free-flow).
  • Absolute Speeds: you can begin calculating delay anytime the speeds drop below a user-defined threshold (example: anytime you drop below 40 mph).

The UDC supplements agency cost analyses related to roadway improvements, impacts of work zones and other conditions.

  • Figure 12: A zoomed in screenshot of user delay costs estimates for stretch of the Capital Beltway in Washington D.C. The exceedingly high UDC costs on January 6th in the morning rush hours were the result of a major snow event.

    Figure 12: A zoomed in screenshot of user delay costs estimates for stretch of the Capital Beltway in Washington D.C. The exceedingly high UDC costs on January 6th in the morning rush hours were the result of a major snow event.

Dashboards

Dashboard allows users to custom build their own personal dashboard of “widgets.” Users can determine which geographies (single corridor from any starting and ending point, multiple corridors, counties, states, etc.) and date ranges (current compared to average, current compared to last year, month-to-date, week-to-date, etc.) are included in their individual widgets. Existing widgets display speed and travel time and bottleneck ranking and have several options for displaying information.

Figure 13: Customized layouts of multiple widgets--defined by each user and tailored to operations and/or planning needs.

Figure 13: Customized layouts of multiple widgets–defined by each user and tailored to operations and/or planning needs.

Figure 14: Zoomed in view of the Corridor Performance Widget shows current speeds and travel times compared to historic conditions.

Figure 14: Zoomed in view of the Corridor Performance Widget shows current speeds and travel times compared to historic conditions.

Figure 15: Zoomed in view of the interactive real-time bottleneck ranking widget which is configurable by certain regions, corridors, metropolitan area, etc.

Figure 15: Zoomed in view of the interactive real-time bottleneck ranking widget which is configurable by certain regions, corridors, metropolitan area, etc.

Future versions of the dashboard will be sharable and/or embeddable within your own agency’s website for quick and easy sharing with the public or other decision makers. When the final MAP-21 rules are released, a version of this dashboard will be made available that is tailored to the MAP-21 reporting requirements.

“Analysis that used to take an entire year to accomplish with one or two full-time employees now takes only 10-minutes, and I don’t need an entire IT staff to support it.”  ~MPO Senior Transportation Analyst

“The VPP Suite represents a quantum leap in capabilities for problem identification, problem confirmation, and communicating with the public.”  ~DOT Planner

“The amount of funding we have to ask for from our DOT program manager has decreased as a result of access to these tools.  They are saving money, and we are more nimble.”  ~DOT Consultant

“We are making better informed decisions about which ops centers to keep open, where to deploy patrols, and what type of economic impact we are having on the traveling public. We’ve never had this type of insight into operations before.”  ~DOT Operations Manager

“Someone finally understands how to display operations data in a way that makes sense and doesn’t make me rely on 10 different systems on 5 different computer terminals.  And it’s fast! I’m simply blown away!”  ~Center Operations Engineer

“This is amazing!  We can tell some really compelling stories to the public about our impact, and it’s so easy!”  ~Private Sector Public Information Officer and Media Relations for a DOT

A screenshot from the "Help" page in the Suite that explains how AADT is mapped to TMC segments for use in delay calculations. All other algorithms are documented as well.

A screenshot from the “Help” page in the Suite that explains how AADT is mapped to TMC segments for use in delay calculations. All other algorithms are documented as well.

Unlike many other analytics tools offered by 3rd parties, the algorithms used behind the VPP Suite are open and transparent. Our algorithms and displays were developed by a mix of academics, professional computer scientists, statisticians, graphic artists, and state and MPO transportation engineers. All of the algorithms used to computer performance metrics are published on-line in our FAQs so that agencies can review them, reproduce them, better understand their meaning, and provide input and feedback for continuous improvement.