EDITOR'S NOTE: Incorrect years were referenced in the interview footage contained within the video above. These have been corrected in the written article.
TUCSON, Ariz. (KGUN) — New data shows Artificial Intelligence (AI) has made a difference at several key intersections across Tucson, where the city has deployed adaptive signalization technology from two companies—NoTraffic and InSync. The goal: reduce congestion and save time for drivers.
Traditionally, traffic signals worked on fixed timing plans, which could lead to unnecessary delays. But with AI technology, traffic signals are continuously adjusting based on real-time data, helping to optimize traffic flow at 92 intersections across Tucson.
Since installing the AI technology in the end of 2021, data from NoTraffic, which utilizes NVIDIA-powered software, shows Tucson drivers have saved more than 1.25 million hours in traffic, cut over $1 million in fuel costs, and reduced red-light running by 80%.
They estimate that intersections using the NoTraffic AI Mobility platform have reduced road delays during rush hours by up to 46% and shortened peak queue lengths by half a mile.
According to the company, NoTraffic’s initial deployment began on Tucson’s West Ajo Way, which led to an average delay reduction of 23% for drivers.
"So (2022), we added some 40 more locations and in (2023), we added more and so every year we added a little bit more," said Francisco Levya, systems engineer for the City of Tucson.
The old model for traffic signals relied on fixed timing plans, often leading to unnecessary delays.
“You would collect some statistics on vehicle counts at the intersection and then decide how much green are you going to give to each direction,” Levya said.
But with new AI-powered software, such as NoTraffic's system running at Broadway and Alvernon, traffic patterns are continuously predicted and adjusted in real time. The lights change based on current traffic flow, optimizing the system minute by minute.
“These new systems are looking at the volume of data second by second, or at least every minute, and they're deciding if there's a change in the traffic pattern,” Levya explained.
When a group of vehicles, or a "platoon," approaches an intersection, the AI can detect if there are any lagging cars at the end of the group.
“The new systems look at the platoon approaching the signal and recognize the end of the platoon," Levya said. "It would extend one, two, or three seconds of that green to allow those vehicles at the end to keep going with the same platoon."
Looking ahead, the city hopes to expand this AI-powered traffic technology to more corridors across Tucson as soon as budget constraints allow.
Levya emphasized the long-term benefits, saying, “If we could do more with what we have instead of adding a lane to a road, if we could improve the capacity of that road that saves Tucson a lot of money in construction and then maintenance afterwards.”
InSync technology is installed at Park and Campbell, from Irvington to Valencia, while NoTraffic's technology is deployed on Campbell, from Limberlost to Ajo, and on Broadway, from Aviation to Swan.
——
Maria Staubs joined KGUN 9 as a multimedia journalist and producer in July of 2024. Her passion for writing and storytelling stems from anchoring her middle school’s news show and editing her high school’s yearbook. She holds a bachelor's degree in journalism and mass communication with a minor in film and media production, as well as a master's degree in mass communication. You can email Maria at maria.staubs@kgun9.com or reach out to her on X/Twitter or Instagram.
