Advertisement

Robots suggest synchronized driverless cars may improve traffic flow

By Brooks Hays
Researchers were able to test the driverless cars in an "egocentric" mode and a "cooperative" mode. Photo by the University of Cambridge
Researchers were able to test the driverless cars in an "egocentric" mode and a "cooperative" mode. Photo by the University of Cambridge

May 20 (UPI) -- New research suggests synchronized, driverless cars can relieve congestion and improve traffic flow by as much as 35 percent.

To test the potential of driverless car technology, researchers at the University of Cambridge built a fleet of small, robotic cars. The cars were released onto a miniaturized track for tests. Researchers measured the impact of a vehicle breakdown on the rest of the fleet.

Advertisement

When the cars were operating independently, the stoppage of a single vehicle caused a significant backup. However, when the cars operated in coordination, communicating with one another as they circled the track, a stoppage in a single lane caused only a brief slowdown.

When the synchronized cars were faced with a breakdown in the inner lane, the robot vehicles in the outer lane slowed down to make it easier for the cars in the inner lane to merge.

Synchronization -- that is, cars in communication with each other -- improved traffic flow by 35 percent. Researchers shared the results of their tests on Monday at the International Conference on Robotics and Automation, held this week in Montreal.

"Autonomous cars could fix a lot of different problems associated with driving in cities, but there needs to be a way for them to work together," Michael He, an undergraduate student at the University of Cambridge, said in a news release.

Advertisement

He helped designed the algorithms that allowed the communicating cars to synchronize their actions in response to a breakdown.

Researchers used motion capture sensors and a Raspberry Pi to allow their miniature robotic cars to sense their surroundings and communicate via wifi. To allow the cars to coordinate their actions, He and his research partners adapted algorithms that real driverless cars use to make safe lane changes.

Similar attempts to test driverless car technologies have either relied on computer simulations or actual automotive vehicles. Using real cars, however, is expensive and requires a lot more space. Researchers at Cambridge suggest their methods could make testing driverless car technologies easier and cheaper. Each of their cars cost about $75 to make.

"Our design allows for a wide range of practical, low-cost experiments to be carried out on autonomous cars," said Cambridge computer scientist Amanda Prorok. "For autonomous cars to be safely used on real roads, we need to know how they will interact with each other to improve safety and traffic flow."

Latest Headlines