With GPS data from as little as 6% of vehicles on the road, University of Michigan researchers can recalibrate traffic signals to significantly reduce congestion and delays at intersections.
In an 18-month pilot study conducted in Birmingham, Michigan, the team used connected vehicle data insights provided by General Motors to check its system, leading to a 20% to 30% decrease within the variety of stops at signalized intersections. GM vehicles make up 6-10% of cars currently on the road in america.
Officially, it is the world’s first large-scale, cloud-based traffic signal retiming system, and it represents a significant opportunity for communities to recalibrate their signal patterns at a reduced cost. U-M’s research appears in Nature Communications.
The U-M system takes GPS data from a percentage of vehicles on the road and extrapolates traffic patterns. For instance, a connected vehicle that involves a stop roughly 100 feet from an intersection strongly indicates that it’s behind no less than three or 4 other vehicles.
“While detectors at intersections can provide traffic count and estimated speed, access to vehicle trajectory information, even at low penetration rates, provides more invaluable data including vehicle delay, variety of stops and route selection,” said Henry Liu, U-M professor of civil engineering and director of each Mcity and the Center for Connected and Automated Transportation.
There are roughly 320,000 traffic signals within the U.S. and the annual congestion costs — direct and indirect — related to those intersections comes out to $22.9 billion. Those costs include time spent waiting at lights, in addition to unnecessary energy consumption attributable to signal times that might be improved.
Most traffic signals operate on a time-of-day signal timing plan, where preset patterns are in place for morning, afternoon, evening and overnight. Traffic planners try to coordinate those cycles with surrounding intersections to permit cars to flow between intersections with as little stop-and-go travel as possible.
“The explanation these signals must be modified more often is that traffic is all the time changing,” Liu said. “A very good example is the traffic patterns we saw within the 12 months before COVID’s arrival and the 2 years afterward. Your morning peak hour modified drastically with so many individuals working from home. Whenever you see that type of change that you must retime these signals.”
Optimizing signals to maintain up with changes in traffic flows is not a walk in the park. The prices and time involved in doing traffic counts and recalculation mean most municipalities won’t reassess for 2 to 5 years, or sometimes a long time.
While adaptive signals have been around for the reason that Nineteen Seventies, detecting vehicles at intersections to reprogram signals almost in real time, cost has kept them from widespread use. Installation of an adaptive system at a single intersection can cost as much as $50,000, with regular maintenance required — a price tag not all communities can afford. The U-M system for optimization would cost a fraction of that for an adaptive system.
The U-M system, called a probabilistic time-space diagram, allows for a smaller percentage of connected vehicle data to do the identical workload as sensors at an adaptive traffic signal. To check its effectiveness, researchers collected data over the course of three weeks in March 2022 from each of Birmingham’s 34 signalized intersections — most of that are fixed-time systems.
“What this has done is basically solve our data collection issue,” said Gary Piotrowicz, deputy managing director of the Road Commission for Oakland County. “And I could argue that that is going to be the way in which everybody within the country does it. Once they’ve solidified the system, there isn’t any reason to do it another way.”
Liu’s team features several graduate students including Zachary Jerome, a graduate research assistant and member of the Michigan Traffic Lab who helped develop U-M’s algorithm. Jerome worked directly with RCOC and hopes to collaborate with industry partners to assist other municipalities deploy this cost-saving technology.
“The chance to work with industry to bring this groundbreaking technology into real-world applications is incredibly inspiring,” Jerome said. “My vision is that this technique will provide a revolutionary signal retiming solution for communities the world over that’s scalable, sustainable and efficient.”
The research was partially funded by the U.S. Department of Transportation and General Motors.
