Inc.
California is testing AI-powered traffic lights to cut congestion without pouring millions into new road construction.
California Just Put AI in Charge of 9 Traffic Lights. It Could Save the State Over $200 Million
Author: Chris Morris
Driving in California is not for the weak. Congestion is common and frustrations pile up quickly. But an ongoing study in the Monterey Peninsula could make commuting less frustrating, assuming artificial intelligence lives up to its potential. At the same time, it could save the state hundreds of millions of dollars.
Officials in Monterey County have turned over the operation of nine traffic signals on the Highway 68 corridor to AI. The goal of the experiment is to see if adding some intelligence to those red and green lights could eliminate the need to build roundabouts along the corridor. And, should that be the case, it could save the area an estimated $200 million in construction costs.
The pilot program for Adaptive Traffic Signal Control (ATSC) officially went live on May 4 (though parts were operational before that), so it’s much too early to gauge success levels. The pilot is slated to last five years. Officials say there are still some bugs to work out, primarily centered around calibration, but the system seems to be working so far.
The AI is overseeing signals in the heavily trafficked area from San Benancio Road to Josselyn Canyon Road along Highway 88. (Highway 88 is the main feeder into the Salinas area from Highway 101.) By monitoring real-time traffic data, the AI is designed to keep traffic moving by adjusting the length of green lights dynamically.
Drivers have been receptive to this use of AI in their commute, saying they’re willing to give it a chance to ease gridlock.
“That doesn’t bother me. It’s everywhere, you know. As long as it helps, I don’t have a problem with that,” Carrie Askey told KSBW.
The savings could be very real. Before the county launched the ATSC program, it had proposed building one-lane roundabouts at intersections (along with, at various points during the debate, widening lanes and building a bypass). Residents, though, pushed back against the proposals as well as the price tag, which reached as high as $215 million.
The AI-driven ATSC system costs just $500,000.
The program was launched just as the school year ends and the summer travel season begins. That was by design. Planners basically decided to throw the AI into the deep end to see how it handled the rush that accompanies summer events in the town.
Additional savings from AI
If it works, the AI-controlled traffic signals could save much more than $200 million. The U.S. Department of Transportation has lobbied for adaptive signal controls for some time. Citing data from the Texas Transportation Institute, it said traffic congestion cost $87.2 billion per year in wasted fuel and lost productivity. That works out to $750 per traveler.
Outdated signals also require maintenance and repairs from state and county DOT workers, which carries a time and financial cost as well.
For commuters, the time savings could be significant, especially when there are unexpected incidents, such as wrecks or event traffic.
“Traditional Time-of-Day signal timing plans do not accommodate variable and unpredictable traffic demands,” the U.S. DOT writes. “Special events, construction, or traffic incidents typically wreak havoc on traffic conditions. While large-scale construction projects and regular events can be anticipated, determining their impact on traffic conditions can be extremely difficult. Other disruptions, such as crashes, are impossible for time-of-day signal timing to accommodate.”
Real-time AI monitoring could fix that. A study of adaptive traffic signal control, released at the end of March, found that at high-demand crossings, AI-driven traffic lights resulted in a 25 percent drop in wait length and a 17.9 percent reduction in accidents.
Credits: TCA, LLC.