Researchers are exploring the use of features in modern cars, such as GPS, to make traffic safer and more efficient. Ultimately, these upgrades could completely replace today's red, yellow, and green lights, handing control over to autonomous vehicles.

Henry Liu, a civil engineering professor at the University of Michigan, led a study that suggests the introduction of a new traffic signal system may be much closer than people realize.

"Artificial intelligence is progressing very rapidly, and I think it's coming," he said.

"When we approach an intersection, if it's a red light, we stop; if it's a green light, we go," said Hajbabaie, whose team used model cars small enough to hold in their hands. "But if a white light comes on, you follow the car in front of you."

Hajbabaie acknowledged that adopting this approach will still take many years, as it requires 40% to 50% of vehicles on the road to be autonomous in order for the method to work.

Sandy Karp, a spokesperson for Waymo, a subsidiary of Google's parent company, pointed out that Waymo has already launched fully automated ride-sharing services in Los Angeles and Austin, Texas, without adding a fourth traffic light.

"While it's good that people are thinking creatively about how to promote the safety of autonomous vehicles in the early stages of their development, policymakers and infrastructure owners should proceed with caution and not prematurely invest in specific infrastructure changes that may be unnecessary," Karp said in an email to the Associated Press.

Researchers at the University of Michigan took a different approach. They conducted a pilot program in the suburb of Birmingham, using speed and location data from General Motors vehicles to change the timing of traffic lights in the city. The researchers recently received funding from the U.S. Department of Transportation, as part of the bipartisan infrastructure law, to test how these changes can be made in real-time.

Since Michigan's research involves vehicles with drivers rather than fully autonomous vehicles, it may be closer to widespread implementation than what Hajbabaie is seeking.

Liu, who leads the Michigan research, said that even with only 6% of vehicles on Birmingham's streets connected to the General Motors system, they can provide enough data to adjust the timing of traffic lights and control traffic flow.

The 34 traffic signals in Birmingham were selected because, like over half of the traffic signals nationwide, they are set on fixed schedules without any cameras or sensors to monitor traffic congestion. Liu said that while there are more high-tech solutions for monitoring traffic, these solutions require complex and expensive upgrades for cities.

"The advantage of this approach is that you don't need to make any changes to the infrastructure," Liu said. "The data is not coming from the infrastructure, it's coming from the car companies."

Danielle Deneau, the traffic safety director for the Road Commission for Oakland County, Michigan, said that the initial data from Birmingham only adjusted the timing of green lights by a few seconds, but it was already enough to reduce congestion. With the new funding, larger changes may be made at a location in the county that has not yet been announced, to achieve automation of traffic lights.