© FreeFoto.com

A system of self-organizing traffic lights could significantly reduce city traffic congestion, according to modeling by SFI External Professor Dirk Helbing of the Swiss Federal Institute of Technology Zurich (ETH Zurich) and Stefan Lämmer of Dresden University of Technology in Germany.

The researchers propose intersections that communicate and make decisions in coordination with others to optimize traffic flow.

City planners today try to synchronize traffic signals to improve flow. But variables like pedestrians, halting trucks and buses, and accidents can cause traffic jams and increase driver idling time and, thus, add to overall fuel consumption and air pollution.

The researchers modeled traffic as if it were a fluid. They first gave each fictional traffic signal smarts that detect traffic at a given moment and make adjustments to green and red time.

But when each traffic light responds only to its immediate demands, the lights just react to the traffic coming from nearby intersections, an approach that does little to improve system-wide traffic flow.

Lämmer and Helbing then tested a decentralized approach that lets the traffic lights communicate and calculate how changes at each intersection would affect the entire system. They found that delay times could be reduced from 10 to 30 percent with system-wide decentralized control.

The researchers are working with traffic authorities in Germany to implement a pilot system based on their approach.

Santa Fe Institute white paper about self-organizing traffic lights

Swiss Federal Institute of Technology Zurich news release

See also: Article in Popular Science

See also: Article in Wired News

See also: Article in Scientific Computing

See also: Article in Money Times

See also: Article in U.S. News & World Report

See also: Article in Discovery News

See also: Syndicated article on UPI.com (United Press International)

See also: Article in CIO.com