North Carolina deploys AI-powered traffic signal management software to 2,500 intersections

The North Carolina Department of Transportation shared Monday it has deployed artificial intelligence-powered traffic signal management software to 2,500 intersections across the state with the hopes of modernizing the state’s traffic operations.

The software, which is from the transportation technology company Flow Labs, has so far been deployed to intersections all across the state. It relies on data from the transportation department’s traffic cameras and GPS data from connected vehicles that traffic controllers can use to improve traffic flow. The data processed using machine learning, then loaded in a singular platform designed to optimize signal timing and improve critical infrastructure planning.

Jatish Patel, CEO of Flow Labs, said his company’s AI makes processing large volumes of data what’s happening on the roadway much simpler, and helps state agencies save money. He said deploying the technology across the state’s infrastructure only weeks, as opposed to the years it may have taken if the system required proprietary hardware.

Philadelphia, for example, is still in the midst of its decades-long process to modernize its traffic control system.

“[We are] able to help them understand how much time vehicles are spending at a delayed traffic signal, or how long the queues are or how much emissions are being wasted in vehicles, just being idled and getting stopped at a signal,” Patel said.
”We also use a lot of artificial intelligence to be able to clean datasets, to be able to find anomalies in data, to clean out bad data so that we can maximize the accuracy of the insights.”

The technology’s deployment across North Carolina, Patel said, is the largest installation of real-time, AI-powered traffic management software in the country.

Patel said Flow Labs recently wrapped a project with the Florida Department of Transportation installing that equipped the central part of the state with software to monitor intersections. He said the traffic data from that system is highly accurate, allowing engineers in Florida to be assured they’re making correct decisions.

“Whenever an engineering team or a state agency wants to get a visibility into what’s happening on the road, they have to deploy a physical device, like a camera, and that can cost up to $75,000 per unit with budget shortfalls. It’s incredibly hard to deploy those as well as maintain those,” Patel said. “So they don’t have the hardware available and often they’re just relying on citizens calling in and complaining about an issue. So they’re really flying blind — they’re not able to really see what’s happening until someone until a problem exists. and it annoys drivers enough that they’re willing to go on the website of the agency, call up and complain about it. The way that we’ve done this is a lot more efficient and a lot more affordable.”