A new traffic signal system designed by University of Michigan researchers is being implemented in Oakland County, with plans to expand across Southeast Michigan. The system uses GPS data to track vehicles as they move through intersections, allowing engineers to regularly adjust signal timing and reflect real-time traffic patterns.
In partnership with the Road Commission for Oakland County, the project uses anonymizedGPS data of roadside assistance vehicles, navigation and ride-hailing services to model traffic patterns across the entire county. The system requires GPS data from as little as 5% of vehicles traveling through each intersection to make accurate projections.
In an interview with The Michigan Daily, Xingmin Wang, engineering postdoctoral research fellow who worked on predicting traffic delays for the project, said the research group is at the forefront of GPS-based approaches to traffic signal timing.
“We are really the pioneers of this direction,” Wang said. “We are one of the first research teams working on this topic, and also the first Ph.D.s working on this topic.”
The system, first trialed in 2022, resulted in a 20% to 30% decrease in the number of unnecessary stops at signalized intersections in Birmingham, Mich.
In an interview with The Daily, Zachary Jerome, postdoctoral research fellow at the Transportation Research Institute, said he believes the early results from the recent rollout confirm the effectiveness the team anticipated after the initial trial in Birmingham.
“They perform about as well as we were expecting,” Jerome said. “One corridor saw up to a 30% reduction in delay, and the other corridor was around 20%.”
Traditional traffic signals rely on fixed-time systems, which use pre-set patterns and require manual updates every few years to adjust to changing traffic patterns. Other systems, such as vehicle detectors and cameras can provide regular traffic data, but are costly to implement and maintain. According to Jerome, the new U-M continuous monitoring system can be scaled widely at a lower cost.
“Oakland County has (more than)1,400 signalized intersections that they manage; to install cameras at all 1,400 of those intersections is not feasible,” Jerome said. “We can use this data that’s being collected continually to know — on a much faster timeframe — what’s happening at each intersection.”
Since its inception, four signals along Eight Mile Road in Farmington Hills and nine signals along Twelve Mile Road in Royal Oak have been updated with the U-M system. Twenty-seven more intersections throughout Oakland County are in the process of being equipped with the technology.
Engineering graduate student Zihao Wang, who worked on the project, said RCOC’s openness made real-world implementation possible.
“I feel very lucky to collaborate with RCOC, because they’re very open minded,” Zihao Wang said. “Our technology is a relatively new thing, and they are willing to try it, and give us an opportunity.”
According to RCOC, Oakland has the most congested roads of any county in Michigan due to its recent population growth, with its population nearly doubling since 1960.
In an interview with The Daily, Craig Bryson, RCOC senior manager of communications and public information, said the road commission aims to explore innovative traffic management methods, a goal the U-M system is helping them achieve.
“We’re always looking for alternative ways to reduce or control congestion, and this provides an incredibly cost-effective, quick, relatively simple way of doing that,” Bryson said. “I hate to use this term, but in this case, it’s really true; it revolutionizes the way we manage traffic.”
Bryson also said reducing unnecessary stops could have benefits beyond improved traffic flow, citing potential impacts on driver behavior from the technology.
“Many accidents at intersections are caused by people running yellow or red lights, and that happens often during rush hour when people are frustrated and tired of waiting at red lights,” Bryson said. “If we can reduce the amount of time those people have to stop at intersections, it just makes sense that would also reduce the amount of times those people are going to take risks.”
Stage one, prototyping and planning, was funded by a $1.45 million Stage One Grant from the U.S. Department of Transportation, supporting the system’s development and rollout at 40 intersections. The U-M research team is currently applying for a Stage Two Grant to enable broader implementation, supported by the newly launched startup Connected Traffic Intelligence.
For RCOC, expanding the system comes down to resources. Looking ahead, Bryson said the system has potential well beyond Oakland County if it receives sufficient access to data and funding.
“Funding is the biggest challenge;access to the data is the next challenge,” Bryson said. “If we could get access to that data at low or no cost, that would make this a no-brainer to expand it county-wide and literally statewide, nationwide, worldwide. Anywhere where there is traffic congestion and traffic signals, this could make a difference.”
Daily News Staff Reporter Niko Wilson can be reached at nikow@umich.edu.
Related articles