Pittsburgh, PA reduces traffic congestion with AI
City utilizes growing network of smart traffic lights to emissions by 21% and journey times by 26%.
Newsworthy
This AI traffic system in Pittsburgh has reduced travel time by 25% | Smart Cities Dive
Smart Cities
This AI traffic system in Pittsburgh has reduced travel time by 25% | Smart Cities Dive
Surtrac Deployment at Urban Grid Networks in Pittsburgh Neighborhoods
Learn about the deployment of Surtrac in the East Liberty neighborhood of Pittsburgh, PA as well as plans to increase the use of the technology at other points in the city.
Surtrac Deployment at Urban Grid Networks in Pittsburgh Neighborhoods
Highlights
Pittsburgh utilizes a growing network of Surtrac smart traffic lights that use AI technology to control traffic signals in real-time.
The system has helped to cut emissions by 21%, journey times by 26%, and wait times at intersections by 41%.
Pittsburgh has received $20 million of funding from federal and state authorities to increase its Surtrac deployment from 50 to 200 installations.
Surtrac is developed by the Robotics Institute at Carnegie Mellon University, which has created the company Rapid Flow to commercialize the technology.
Summary
Back in 2012, the City of Pittsburgh first implemented the Scalable Urban Traffic Control (Surtrac) system to coordinate its traffic lights in real-time. The system is developed by the Robotics Institute at Carnegie Mellon University, and it takes advantage of artificial intelligence and predictive data to adapt traffic light sequencing plans at each individual intersection in real-time.
Traditional signaling systems operate on set timing plans that cannot change in response to actual traffic flow and evolving traffic patterns. With Surtrac, each light builds its own traffic plan by gathering data on approaching traffic. Rather than tracking individual vehicles, the system aggregates traffic into groups and maps its position relative to other groups. This data enables Surtrac to predict the size of traffic clusters, when they will arrive at the light, and when they will clear the intersection. Each light communicates the data it gathers to neighboring lights, which makes it possible for the system to build long-horizon plans and accurately adapt sequencing to minimize traffic buildup.
Initially funded by the Hellman Foundation, the project began with 9 intersections. By 2016, Pittsburgh had scaled out to 50 installations; and it has received more than $20 million of funding from federal and state authorities to add another 150 installations over the next few years. As Surtrac is a decentralized system, intersections can easily be added to the network over time.
$ Rapid Flow$ – the company created by Carnegie Mellon University researchers to commercialize Surtrac – reports that the system has helped reduce travel times by 26%, emissions by 21%, number of stops by 31%, and wait times at intersections by 41%.