[Note: This item comes from friend Mike Cheponis. DLH]
Pittsburgh’s new artificially intelligent stoplights could mean no more pointless idling
By Chris Weller
Oct 22 2016
Traffic lights are finally getting smarter in Pittsburgh.
Thanks to a new pilot program from the tech startup Rapid Flow Technologies, Steel City now boasts 50 intersections whose stoplights are running artificial intelligence software known as Surtrac that reduces wait times on empty or lightly-traveled roads.
Since Surtrac was first introduced in 2012, the Rapid Flow team estimates the AI stoplights have cut emissions by 21%, travel times by 25%, and idling times by 40%.
The magic of Surtrac is that it bundles each stoplight into an intelligent network “that moves all the vehicles it knows about through the intersection in the most efficient way possible,” Rapid Flow CEO Steve Smith said at the recent White House Frontiers Conference, according to IEEE Spectrum.
Surtrac relies on a system of cameras and radar sensors that detect traffic patterns in particular areas. When one area starts to see more traffic — during rush hour, for example — the other stoplights use a proprietary set of algorithms to adjust their timing accordingly.
The result is a smarter city that operates more like a living, breathing organism than just a static patch of roads.
Pittsburgh has recently been a popular site for urban-planning innovation. In August, the city played host to Uber’s first rollout of self-driving cars. Uber selected Pittsburgh because of its odd assortment of narrow, one-way streets mixed with steep hills and a staggering 446 bridges, all of which make it an ideal setting for testing the limits of AI.
As IEEE Spectrum reports, Surtrac isn’t the only smart traffic-management system. There are others in Utah, California, and Washington. But unlike those systems, the stoplights in Pittsburgh don’t need a jumble of wires run beneath the city streets or the help of a central command to run.
In Pittsburgh, Surtrac allows the lights to talk to one another independently, based only on the feedback from the sensors and cameras. They essentially think for themselves.