The Waiting Is (No Longer) the Hardest Part

“My citizens get [upset] when they see [poor] signal timing,” said David Hansen, a 30-year Army veteran and city manager of the City of Virginia Beach, in his opening remarks at the Virginia Section of the Institute of Transportation Engineers (VASITE) Annual Meeting in June 2017. Using more colorful words, Mr. Hansen went on to capture what agencies managing signal systems hear from their customers on a daily basis: It’s not just having to wait, it’s having to wait when you shouldn’t.

Fortunately, the transportation profession has a better way to address this challenge using Automated Traffic Signal Performance Measures (ATSPMs). ATSPMs combine high-resolution (tenth-of-a-second) data from traffic signal controllers and detectors (real data!) along with data analysis techniques. The concept gained traction in 2012 through the collaboration of innovators and pioneers in the transportation industry such as Jim Sturdevant of the Indiana Department of Transportation (INDOT) and Darcy Bullock of Purdue University. This team, in partnership with other state and local departments of transportation, the Federal Highway Administration (FHWA), and signal control vendors, developed a standard for how to log event data regardless of hardware or software type. The development and buy-in of this standard opened the door for widespread application.

Then along came the Utah Department of Transportation (UDOT). Inspired by senior leaders to be “world class,” UDOT developed an in-house automated system for displaying signal performance measures in 2013. By 2016, UDOT had updated to its fourth version, which it made available to all other transportation agencies through the OSADP open-source software portal maintained by the USDOT. The UDOT ATSPM software is currently available online at https://github.com/udotdevelopment/ATSPM.

ATSPMs are gaining national attention partially due to FHWA’s action to include ATSPMs in its Every Day Counts program. As a result, many more agencies are installing the UDOT ATSPM software, and even vendors are beginning to incorporate the metrics into their systems. Every Day Counts is a state-based model launched in 2009 that identifies and rapidly deploys proven-but-underutilized innovations to shorten the project delivery process, enhance roadway safety, reduce congestion, and improve environmental sustainability.

But this is just the beginning.

Together, Kittelson and Purdue are developing guidance on the Performance-Based Management of Traffic Signals, which will describe signal performance measures, how they are derived, and how they can be applied to improve the performance of signal systems. The guidebook, due to be completed in spring 2018, will outline for agencies and practitioners how to select, implement, and apply ATSPMs to a signal system – large or small.

Grant Sanders, who manages over 400 traffic signals for the Virginia Department of Transportation’s Northwest Region, attended a workshop hosted by UDOT in January 2015 and was intrigued by its capabilities, so he decided to try it out for an 18-intersection corridor along US 29 in Charlottesville, Virginia. The corridor carries 55,000 vehicles a day and is a heavily accessed commercial corridor undergoing major construction. Kittelson began working with VDOT to upgrade to the latest UDOT ATSPM software in December 2016. By spring 2017, Grant and his team were using the data to respond to public service requests and make improvements to the signal timing.

In one example, during an evening with back-to-back soccer practices, travelers called VDOT because a side street was overwhelmed with traffic from both pick-ups from an earlier practice and drop-offs for a later practice. Within a matter of minutes, Grant and his team were able to verify and diagnose the problem (something not previously possible without sending a technician to the field). The problem was solved by simply adjusting a timing parameter that gave unused green time to the side street in need. The phone calls stopped, and the before/after impacts were immediately recorded.

In a second example, new timing plans had been implemented by a contractor performing construction along the US 29 corridor. Grant and his team manually fine-tuned the offsets for the mid-day plans. To verify that the changes made an impact, they turned to the Purdue Link Pivot Report (supplementing with Purdue Coordination Diagrams).

The Purdue Link Pivot Report shows the power (and simplicity) of data analytic techniques. Using the actual arrival information measured in the field, this report provides information at the corridor level versus for individual intersections. Reports for before and after conditions showed a 10 percent increase in the amount of traffic arriving on green during the mid-day plan.

Using BlueMAC units installed along the corridor, the team was further able to confirm the impact of the mid-day offset adjustments. Based on the Purdue Link Pivot Report, the southbound direction gained the most from the signal timing changes. Looking at average speeds in the southbound direction tells the same story in a different way. Because more vehicles arrive during green with the new offsets, speeds have increased during the mid-day to better align with corridor trends.

These are just a few of the stories that are beginning to be told. Let us know yours, and how you think ATSPMs can help move our profession forward.