Traffic Lights Are Getting Even Smarter and More Connected

A project in Pittsburgh using super-smart traffic lights is easing gridlock as well as emissions.

By on February 18, 2015

Those of us in urban and suburban environments might remember that fateful day when proximity sensors were added to traffic lights. Suddenly, the lights could think for themselves -- no more sitting at a stoplight in the middle of the night, pounding on the steering wheel, waiting for the light to change when there’s nobody coming the other way, because now the lights can tell when there’s a need for them to change. “There are technologies that can detect when a car has entered a turn lane, so that they can activate the arrow light, or extend the duration of a green light after detecting that there is an increased number of vehicles at an intersection, or activate the green light when a car is approaching an intersection,” says Jordan Perch, an analyst at DMV.com. It’s a small thing, but it felt like magic. Now a new generation of smart traffic lights are taking that magic to a whole different level.

The first major breakthrough in smart traffic lights came in 2012, thanks to the Traffic21 initiative at Carnegie Mellon University in Pittsburgh, PA. Pittsburgh is a particularly difficult city to monitor, traffic-wise; entire blogs have been set up to enumerate the insanities of driving in the city. (Some examples: a three-way intersection where literally nobody has the right of way, bike lanes that totally ignore main arteries of the city and stop signs in the middle of highways --  not to mention the weirdness of having roads intersect with themselves.) Pittsburgh clearly needed some help in the planning department, and while traffic lights may not seem an obvious solution, they certainly provided one -- or at least the beginnings of one.

Carnegie Mellon’s Traffic21 project actually uses artificial intelligence to allow connected traffic lights to talk to each other and help reduce traffic where it’s heaviest. Using proximity sensors, visual sensors and audio sensors, they can figure out the minute-by-minute traffic patterns in the confusing city and adjust their timing and rhythm to help ease that traffic. And the system works: it resulted in 40% less waiting time at stop lights, which in turn helped lead to a projected 21% reduction in vehicle emissions, because the less time you spend driving and, especially, the less time you spend idling, the fewer emissions your car gives off.

On the other hand, these kinds of technologies can make people impatient and no longer used to waiting while driving. “These systems could make drivers less patient in the future, since they will get used to driving through intersections without having to stop at red lights and wait for a few minutes before they can proceed,” says Perch. But that seems a small price to pay for enabling us to spend less time in the car, polluting the air with emissions.

Tags: Traffic21, Carnegie Mellon, Pittsburgh, smart traffic lights, IoT