Moving to end the persistent problem of red light running

by Merry Rendahl

Safety is the major concern of engineers, and we all rely on engineers to make and keep our world safe. Researchers in CEGE have partnered with MnDOT to solve one of the most persistent safety problems in transportation engineering. The problem involves not only built infrastructure but also the behavior of human drivers, drivers who run red lights.

Reducing or ending crashes due to red light running would save lives, save money (from damage done to individual property and city infrastructure), and reduce lane closures and back-ups due to a crash. As desirable as a solution would be, it has proven difficult to change driver behavior.

Currently, the traffic signals are timed in a standard way and all drivers see the light change from green to yellow to red at the same time. A driver turning onto a road might see a yellow light, but would have no information about how long the light had been yellow or how soon it might change to red.

Previous research by Gary Davis (professor emeritus) showed that often drivers do not slow down for a yellow light, but speed up. Presumably to get through the intersection and avoid waiting through a red cycle. 

Professor Michael Levin’s research group has been working on ending red light running. Their idea is to give drivers more information about the intersection they are approaching and deliver it in a more timely way. The enhanced information system holds promise for solving the persistent red light running problem, and the idea has support from the Minnesota Department of Transportation (MnDOT). This system is primarily useful for inattentive drivers or drivers who misjudge their braking capability.

Graduate student Maziar Zamanpour (advised by Michael Levin) developed a computer program called the Red Light Warning System (RLWS) that combines data about the intersection including geometry and timing cycles on the traffic lights, with real-time data from vehicles, including geospatial location, speed, and direction. The data is delivered to vehicles in real time to give drivers enhanced warning information. 

Last summer the team installed the program on one traffic light in Shakopee, Minnesota. Researchers pilot tested the system. The results were positive, so this summer they were able to set up an expanded test drive loop using five intersections, civilian drivers, and real traffic. 

The system set-up requires sensors and transponders on the traffic lights and in each vehicle. RLWS combines data from vehicles with traffic signal timing and intersection geography—more data than traffic lights alone or GPS systems alone. It delivers the enhanced information to give drivers personalized data based on their driving. The aim is to empower drivers to avoid a crash or unnecessary traffic disruption due to running a red light.

Setting up test runs for this type of technology can be challenging and involves some risk. Before these researchers got to the road test stage, they spent two-and-a-half years developing and testing the computer model and demonstrating that the model worked. They first worked with MnDOT to install the technology on one traffic light in Shakopee so they could test the software model with one traffic light with trained researchers as drivers. That first test went well and proved the concept of the RLWS system could work. 

After rigorous testing and safety demonstrations, MnDOT agreed to expand the testing. The technology (sensors and transponders) were installed at five traffic lights in a connected loop in Shakopee for Phase 2, testing with civilian drivers in live traffic. Although some car companies had tested similar programs on their proprietary closed road systems, the five connected intersections made RLWS Phase 2 the largest live road test of its kind. 

The Test Drive

On a sunny Sunday afternoon in July, the researchers were ready to test the system on the live road loop with citizen drivers. This was a rare opportunity for the graduate students involved to get experience with a live road test.

I, too, had a rare opportunity to participate in this live driving experiment. When it was time for my test drive, Maziar Zamanpour set the system up in my car. It consisted of a sensor that could “talk” to the transponder on the traffic lights and a feedback screen mounted to my windshield, where it was easily visible while driving. That screen was the only thing I needed to pay attention to in using the RLWS. 

The researchers, who rode along on all the test drives, were also attending to laptop computers where they could view the real time data coming and going between the vehicle and the traffic signals. I was riding with Mazier, who had written the program. He was attending to a lot of data pouring in. He was also attending to the route and let me know when to turn so I would pass through all five of the connected intersections. I drove the loop twice.

As I started driving, the feedback screen displayed a green circle, indicating that my driving was appropriate for the situation. As I approached the first intersection, I saw the feedback screen display a yellow circle, indicating that I should begin to slow down. Then a red circle appeared indicating a need to apply the brakes more firmly, which I did, and came to a safe stop. For a driver approaching the intersection at a speed that might prevent a safe stop, the red circle would swell with urgency (“hit the brakes, now!”). 

As I approached another intersection and saw the signal was yellow, I would have been inclined to begin to stop, but the RLWS showed a green circle indicating that I could continue at my current speed. Sure enough, the signal changed to green as I approached so no slowing or stopping was required. Such guidance can ensure a smoother flow of traffic.

In my experience, the RLWS worked as planned. It was not intrusive for me as a driver, and did seem to make my drive smoother with fewer stops and starts. The goal is that the RLWS will reduce the numbers of drivers who unsafely enter an intersection as the light turns red.

My approach to driving that day was calm and cautious. However, I cannot say I never rush a yellow light or never feel rushed or eager to get home. In those moments, I seem to fear waiting through a red-light cycle more than I fear a crash. The RLWS could be a good encouragement in those moments. And I suspect some drivers feel even more rushed than I do.

So, until the new technology is fully integrated into our transportation system, drive carefully. Patient and attentive drivers can act right now to prevent crashes due to red light running.