Can AI put an end to traffic jams?

During rush hour, cities around the world can grind to a halt. Yet even as vehicles, cyclists and pedestrians crowd intersections, many traffic lights continue to operate according to fixed timing plans rather than the conditions unfolding on the road.

Israeli startup NoTraffic, founded in 2017, believes artificial intelligence could fundamentally change that.

The company has developed what it describes as an autonomous urban traffic-management platform designed to monitor intersections in real time and adjust traffic signals within milliseconds. Its ambition is straightforward: reduce congestion, improve safety and make urban roads more efficient.

“Traffic has gone relatively through very little change in the past 100 years,” a company executive explained. “We thought that using technology, we can make a major breakthrough.”

The idea behind NoTraffic's technology is simple. Instead of relying primarily on statistical predictions about when traffic is likely to arrive, the system continuously observes what is actually happening at an intersection.

Smart devices installed at traffic lights combine cameras, radar and AI processing to detect and classify road users, including cars, buses, cyclists and pedestrians. The system can then adjust signal timing based on the current flow of traffic.

Consider a driver arriving at an empty intersection at 2 a.m. Under a conventional traffic-light schedule, that driver might be forced to wait at a red light despite being the only vehicle on the road.

With a real-time system, the light could recognize that there is no conflicting traffic and change accordingly.

It is a small example, but one that illustrates the broader principle: traffic signals should respond to traffic rather than simply follow a clock.

NoTraffic says its platform can also coordinate traffic across entire corridors rather than treating intersections as isolated points.

In one example presented by the company, traffic around a major U.S. football stadium was managed during a Monday-night game without relying on police officers to manually direct vehicles. According to NoTraffic, the optimized corridor performed better than it would have under traditional traffic management.

The company's demonstrations show a similar effect. In one comparison, traffic conditions at an intersection were recorded before the system was installed and then again two weeks later at the same time and on the same day of the week.

The difference was striking: a large queue of vehicles had disappeared, while more cars were moving through the intersection. The system also detected problems such as a backed-up left-turn lane and adjusted signal timing to clear it.

According to NoTraffic, deployments in cities have reduced congestion by 30% to 50% in many cases.

NoTraffic estimates that for every dollar cities spend on its technology, they can receive between $75 and $200 in economic benefits. The calculation, according to the company, includes savings associated with reduced congestion, lower emissions and improved road safety.

Fewer unnecessary stops can mean less fuel burned and fewer emissions. Better coordination can also reduce dangerous situations, including red-light violations and collisions.

For municipalities facing pressure to improve transportation without endlessly expanding roads, technology offers another possibility: making existing infrastructure work better.

Most cities already have traffic-detection systems in place, ranging from conventional sensors to video, radar, thermal and LiDAR technologies. NoTraffic's approach is to use and upgrade this technological infrastructure rather than requiring cities to rebuild their entire road networks.

The company is also looking beyond today's traffic.

As autonomous vehicles become more common, vehicles will increasingly be able to communicate with one another and make their own decisions. But NoTraffic argues that intersections will still need an overarching system capable of understanding the bigger picture.

The company says its deployments include dedicated communications hardware designed with future connected and autonomous vehicles in mind.

The broader vision is a city in which vehicles and infrastructure continuously exchange information, allowing traffic to be coordinated dynamically rather than controlled through static rules.

Artificial intelligence has not eliminated traffic jams. But if NoTraffic's technology delivers the results it claims, the future of urban transportation may look less like a network of isolated traffic lights—and more like one intelligent system responding to the city in real time.