Researchers argue for smarter traffic lights

Traffic light-based analytics could be used to help speed traffic, Santa Fe Institute researchers argue

Santa Fe Institute researchers devise a superior way of operating traffic lights so traffic moves more quickly

Those routinely stuck in urban traffic jams can take heart in the fact that scientists may have found a way to reduce congestion, by rethinking the way traffic lights should operate.

Two European researchers, Dirk Helbing of the Swiss Federal Institute of Technology Zurich, and Stefan Lämmer of Dresden University of Technology in Germany, propose changing the way intersection traffic lights are timed, using a combination of sensing technology, analytics and networking.

Rather than the usual approach of coordinating the timing the lights along the road in a way that anticipates the usual flow of traffic, the researchers suggest letting traffic lights themselves judge when to turn green or red.

"Instead of waiting for a certain point in time before switching to green, we now wait for a critical number of vehicles ready for service at maximum rate which given by the saturation flow," they argue in a Santa Fe Institute study

Such an approach could reduce congestion by as much as 10 to 30 percent, the researchers claim.

On most heavily trafficked roads, traffic lights along the roads are typically timed and coordinated using a top-down approach. Traffic engineers schedule the timing of the lights to anticipate the usual traffic patterns. Multiple traffic lights can be coordinated to allow a cluster of vehicles to travel along the road without stopping, an effect traffic engineers call a "green wave."

While this approach helps immeasurably in speeding cars to their destination, the researchers argue that it still doesn't take into account the full variations of traffic.

Traffic rarely flows in a manner as smoothly as traffic models suggest, they suggest. Accidents, buses making stops, overly aggressive or timid drivers, and pedestrians crossing the road all can further distort the modeled traffic patterns.

"It is actually not optimal control, because that average situation never occurs," Helbing told Science News. As a result, even with the best-timed traffic management systems, drivers to sit through multiple red lights, or unduly wait behind a red light when there is no opposing traffic.

The researchers propose giving each traffic light some autonomy in choosing when to change. A traffic light could be connected with sensors that show upcoming and leaving traffic. With some internal processing, the light then can judge which road has the greatest need for the green light.

"In contrast to a fixed-time controller ... the green times are requested only when there is definite demand for them," the researchers write. "The cycle time is not fixed, and the service is not necessarily periodic."

At least one IT systems vendor, IBM, sees a market opportunity in more advanced traffic management systems.

IBM Research is developing an urban traffic prediction system, called the Traffic Prediction Tool (TPT). This system, being tested in a number of different cities such as Singapore, can take input from various road sensors and, using a traffic flow model for that city, not only show where the traffic jams are now, but even can predict where traffic jams may occur. With this knowledge, traffic management departments can make adjustments of the traffic, through road signs that suggest alternate routes.

"We calibrate a set of models on the most recent data, and in real time these models are applied to the real-time data feed," said IBM researcher Laura Wynter, who works on the system.

While giving the traffic lights room to make decisions is a cornerstone of the Santa Fe proposal, the researchers admit that giving full autonomy to each light would produce chaos for the traffic flow as a whole. The traffic light would still need some outside guidance. They propose a method of dynamic control where each light could factor in the traffic patterns of its neighboring lights, and together they could produce the most efficient traffic flow overall.

"Indeed, in order to achieve the ideal 'green wave' in a signalized traffic network, it is necessary to have predictions of the near-term traffic approaching the signals," said Wynter about the study.

Joab Jackson covers enterprise software and general technology breaking news for The IDG News Service. Follow Joab on Twitter at @Joab_Jackson. Joab's e-mail address is Joab_Jackson@idg.com

Insider Tip: 12 easy ways to tune your Wi-Fi network
Join the discussion
Be the first to comment on this article. Our Commenting Policies