Cutting Congestion With Talking Cars

In Singapore, MIT researchers are experimenting with a system that allows cars to communicate to prevent a traffic jam.

Cutting Congestion With Talking Cars
[Top photo: 1000 Words via Shutterstock]

Reducing congestion costs money. When London rolled out its “congestion charge” scheme in 2003–a system of cameras and sensors ringing the city center that charged drivers at peak traffic times–the price was more than $260 million. Though it eventually saw benefits in faster roads and better transit, the bill was initially a bitter one.


Well, in the future, congestion pricing may not be as expensive as that, and it may be more flexible as well. A new system being developed by MIT researchers would do away with expensive infrastructure in favor of a phone-based peer-to-peer network between cars. And, it would also allow planners to move congestion zones as they wished–say, to accommodate an event or an emergency.

Jason Gao and his PhD adviser Li-Shiuan Peh developed the RoadRunner system for the Singapore government as part of MIT’s Future Urban Mobility project. It’s based on two pieces of hardware: a wireless radio on every dashboard (using a form of Wi-Fi) and an app on everyone’s phone. When a city decides to limit how many cars can enter a zone, it issues a certain amount of tokens, like tickets to get into a club. Thereafter, a car can enter only if they get a token from another driver who’s leaving the zone; everyone else is redirected away from the area with GPS-like traffic directions.

Flickr user jeremyfoo

Two things are novel about the system. First, it does away with physical infrastructure at the roadside. “Normally, if [cities] want to change their regions over time, they have to continue constructing new physical infrastructure,” Gao says. “[With this system] because you don’t have to go through toll booths or under gantries, you can electronically redefine areas.”

The other thing is the system works driver-to-driver. When you need a token, your car will contact the next available car; it will only contact the central server if there are no other cars offering tokens. That potentially makes the system quicker than a car-to-server type set-up, and potentially improves privacy as well.

Gao and Peh did two experiment to test the software. First, they put radios and phones in 10 cars and drove them around Boston. That showed how the communication was speedier than a car-to-server network.

Second, they ran a big simulation using traffic data from the Singapore Land Transport Authority. The test improved peak hour traffic flows by 7% compared to the city-state’s existing electronic pricing system.


Gao doesn’t know when or if someone might install the system for real. The project is just for now. But it may not be too far away. There’s not much point paying $260 million if you can have a better system for less than that.

About the author

Ben Schiller is a New York staff writer for Fast Company. Previously, he edited a European management magazine and was a reporter in San Francisco, Prague, and Brussels.