Feature | Susan Langdon, P.E., PTOE | February 08, 2017

Signaling a change of way

Traffic signal engineers can now directly measure what they could previously only estimate and model

There are approximately 400,000 signalized intersections in the U.S. Delays at traffic signals contribute 5–10% of all traffic delays on major roadways alone. What if we had a systematic procedure for identifying traffic operation problems—and fixing them using controller and probe data? Well, we do, and it is something Purdue University, the Indiana Department of Transportation (INDOT) and the Utah Department of Transportation (UDOT), along with several professional organizations, are making a concerted effort to implement. But first, traffic engineers and planners must educate the general transportation community about new automated traffic signal performance measures that can change the way DOTs operate.

 

For decades, the practice of signal-timing management has largely gone unchanged. Traffic-signal timing plans are developed from volume data collected at intersections for a few hours on a typical day; timing plans can remain in place for three to five years or more before they are reevaluated. What’s the problem? The plans are not always effective because they become outdated when traffic conditions change due to growth or other factors. Thankfully, several bright minds set out to solve the problem and developed a dynamic solution.

 

Today, we have a better way, a new approach that is one of the most significant improvements in signal-timing management in decades. Dr. Darcy Bullock of Purdue University, in collaboration with INDOT and UDOT, has combined research efforts to develop a new traffic and signal management approach, known as Automated Traffic Signal Performance Measures (ATSPMs). These measures allow technical professionals to analyze intersection operations like never before by using high-resolution data. It has been a long process that required many organizations coming together to institute change, starting with the “Pooled Fund” study.

 

 

Some background

In 2011, 12 organizations collaborated on the Federal Highway Administration (FHWA) Pooled Fund study. The study TPF-5(258) was led by INDOT and was supported by nine other state transportation agencies, the FWHA Office of Operations and Resource Center, and the Chicago Department of Transportation (CDOT).

 

The study focused on developing automated traffic signal performance measures (ATSPMs) based on high-resolution controller event data (from here referenced as event data). Event data encompass vehicle detection information and traffic signal phase information, recorded at 0.1-second time intervals. The ATSPMs use these data to allow agencies to visualize traffic being served at intersections and along corridors at all times of the day. These tools allow agencies to detect issues more quickly, measure travel times along corridors, and optimize signal operations like never before.

 

Why is this event data different than the data collected from previous methods? This new high-resolution data contains a real-time log of just about every “event” the traffic-signal controller records. These events include changes in detector states (on or off), signal phases (begin green, begin yellow, begin red), and all other actions the equipment controls and monitors. The idea to record this high-resolution data came from Purdue University, which originally partnered with INDOT and signal controller manufacturers to develop a data-logging system that could be used across the board by all manufacturers. Once established, Purdue took data-logging information to another level when they developed a series of performance measures. INDOT tested the data and discovered the value of the information provided. The data offered INDOT staff a modern new way to adjust and monitor their traffic-signal system. They built a website that graphed the ATSPMs to improve the ease of use for their team.

 

UDOT was introduced to SPMs in 2012 by Purdue and INDOT. Mark Taylor, traffic signal operations engineer at UDOT said, “In 2012, UDOT’s executive leaders wondered what it would take for UDOT’s traffic signal operations to be world-class. We knew we needed to implement real-time monitoring of the system.” So they used the performance measures created by Purdue and INDOT. Their staff followed INDOT’s lead and developed software to digest the data and customize the reports for their needs.

 

Figure 1. The Purdue Coordination Diagram.

 

What do the automated performance measures do?

The new performance measures use high-resolution data recorded at 0.1 per second intervals, which are stored in the signal controllers. The information from each intersection is downloaded to a database server. Software on the server converts the recorded data into meaningful information. The performance measures are then displayed in easy-to-understand graphics. This real-time, continuous information provides agencies a much more detailed and accurate understanding of their intersection and corridor operations. Some of the metrics currently available include:

  • Purdue coordination diagrams (PCDs);
  • Approach volumes;
  • Approach speeds;
  • Purdue phase termination charts;
  • Split monitor;
  • Approach delay;
  • Arrivals on red;
  • Pedestrian delay;
  • Preemption details;
  • Turning movement counts; and
  • Purdue split failure.

 

Some of these metrics, such volumes, speed and delays, are familiar to most traffic engineers. But there are some exciting, new metrics that provide a succinct visual summary of how coordinated signals are operating. The Purdue Coordination Diagram (PCD) (Figure 1) shows a 24-hour timeframe for an individual signal phase at the intersection, identifying when the phase was green, yellow or red. It also shows each actuation (vehicle arrival) throughout the day represented by a dot on the graphic identifying whether the car arrived during the green, yellow or red phase. Most traffic engineers, when working on signal timing, spend hours in the field watching the signal operate. It is time-consuming, and it does not always depict the full picture. The PCD shows an entire day of operation—what is working well and what needs adjustment—all from the ubiquitous vantage of the office. The software even provides daily email alerts.

 

UDOT’s software has moved the agency a long way down the road towards offering a “world-class traffic signals” program. UDOT is so invested in the implications its software may have on the national transportation community, it has made its software available to the public for free. Taylor notes, “UDOT created a website that does not require any special software or passwords to access. In fact, you can pull this website up on your smartphone. This allows anyone who has a desire for the data to use it. We have approximately 1,700 intersections statewide in Utah bringing back near real-time SPM data. Using these ATSPMs, UDOT has been able to improve mobility, increase safety and use resources more effectively. We now find most traffic-signal problems before the public notices.”

 

Centralized traffic management software developers are also taking note of the benefits of using ATSPMs. Many “off-the-shelf” systems that are under development or available now offer many of these features in their new systems. The software has become so beneficial that the FHWA made it available on its website for open-source software. The federal agency also included software documentation and a venue for users to share improvements in an open forum for online discussion. 

 

According to Dr. Bullock, the combination of agencies and strong vendor participation was critical to the success of this research. “Agencies reached consensus on performance measures that help them more effectively manage their traffic signal assets. Vendors were quick to pitch in and deploy the requisite high-resolution data collection capabilities needed to compute the new performance measures,” says Bullock. “It’s truly been a collaboration of all the various user groups that makes up the transportation community to implement these new practices. It’s an exciting time in the history of traffic-signal management.”

 

Figure 2. Tracking actions taken and ATSPM tool employed to correct.

 

Why use automated measures?

What can an agency accomplish using ATSPMs?

 

There are many opportunities in these measures to improve daily operation of signal systems and the long-term health of the overall signal network, including:

  • Troubleshoot complaints and reduce wasted time for maintenance staff;
  • Identify problems more quickly, without waiting for a complaint call;
  • Operate and optimize system more efficiently;
  • Retime signals as needed, not on a schedule;
  • Communicate system performance to public agency leaders;
  • Transition from reactive management to proactive signal management; and
  • Improve overall safety.

 

Everyone has a vested interest in reducing congestion, saving fuel costs and improving safety—from transportation departments and metropolitan planning organizations, to municipalities and other agencies seeking to improve mobility in their community. By maintaining safe and efficient signal timing, DOTs can relieve congestion and help people get home safely.

 

In the past, UDOT tracked how often staff took action based on ATSPMs and how these actions improved both operations and maintenance. Figure 2 reflects the number of actions taken to correct different problems and the ATSPM tool that was used to find those problems between Jan. 1, 2014 and June 18, 2014.

 

Figure 3. A diagram of agencies employing ATSPMs.

 

“You can see that we identified and fixed more than 250 detector problems, used the data more than 200 times in our traffic signal models, made more than 150 signal coordination split adjustments, and corrected or optimized more than 50 time-of-day schedules for signal coordination, Taylor said. “These ATSPM’s have enabled us to do more with less, focus our resources on the areas that need it most, and allowed us to more effectively prioritize resources and workload.” That’s why Taylor is not surprised to see that transportation departments across the country are now starting to realize the benefits and efficiencies the ATSPM software can deliver.

 

Currently, the city of Richardson, Texas, is working alongside professional consultant Stantec to set up the software for ATSPMs to manage its signal system. The software is not something you buy off the shelf. Having a traffic-signal consultant onboard can help agencies understand the software, coordinate installation and train staff on best practices. The ATSPMs are another tool DOT and municipal staff can use to provide them with better-quality information in traffic-signal monitoring to make their jobs easier. 

 

The FHWA’s Every Day Counts program includes ATSPMs as one of 11 innovations featured in the Phase 4 2017-2019 program. The mission of the EDC-4 initiative is to accelerate the implementation of automated traffic-signal performance measures by providing tools, technical assistance, and outreach to advance objectives and performance-based approaches to traffic-signal management and operation. The UDOT software is available on the FHWA Open Source Application Development Portal (www.itsforge.com), including documentation. The software’s availability is made possible thanks to collaboration between the FHWA and UDOT, which will hopefully lead to increased traffic monitoring throughout the nation.

 

So, with all this great opportunity, where can DOTs go from here? Agencies can reach out to their local FHWA support team for information on funding opportunities to enhance their signal system and incorporate ATSPMs into their agencies. They can contact agencies that have already begun using ATSPMs (see Figure 3) to understand lessons learned and discover how they made use of the information from a local perspective. Moreover, the FHWA website provides discussion boards on ATSPMs, and there are webinars and professional society meetings within the industry for further information, as well as professional consultant teams who offer assistance in making ATSPMs a reality for any community.

 

Overlay Init