From guardrails to potholes, AI is becoming the new eyes on America’s roads.

As America’s aging roads fall further behind in much-needed repairs, cities and states are turning to artificial intelligence to identify the worst hazards and decide which fixes should be prioritized.

For example, Hawaii officials are giving away 1,000 dashboard cameras in an effort to reverse the recent spike in traffic fatalities. The cameras use AI to automate the inspection of guardrails, road signs, and pavement markings, instantly identifying minor issues and emergencies that require dispatch of maintenance personnel.

“This isn’t something you look at once a month and then sit down and figure out where to put the van,” said Richard Browning, chief commercial officer at NextBase, which developed the in-vehicle camera and imaging platform for Hawaii.

After San Jose, California, began installing cameras on street sweepers, city officials confirmed that the system accurately identified potholes 97% of the time. Now they’re expanding their efforts to parking enforcement vehicles as well.

Texas, which has more road miles than its two neighboring states combined, is less than a year into a major AI initiative that uses camera and cell phone data from registered drivers to improve safety.

Other states use the technology to inspect road signs and produce annual reports on road congestion.

All guardrails, every day

Over the next few weeks, Hawaii drivers will be able to sign up for a free dash cam valued at $499 under the Eyes on the Road campaign. The campaign was piloted on service vehicles until it was suspended in 2021 due to bushfires.

Roger Chen, an associate professor of engineering at the University of Hawaii who is helping promote the program, said the state faces unique challenges in maintaining outdated road infrastructure.

“We need to transport the equipment to the island,” Chen said. “This is not a simple problem because you have to deal with space constraints and terrain constraints.”

The program also monitors things like street debris and faded paint on lane markings, but the company behind the technology specifically touts its ability to detect damaged guardrails.

“They analyze every guardrail in the state every day,” said Mark Pittman, CEO of Brinsea, which uses a combination of dashboard feeds and mapping software to analyze road conditions.

Hawaii transportation officials are well aware of the risks posed by broken guardrails. Last year, the state reached a $3.9 million settlement with the family of a driver who died in 2020 when he struck a guardrail and left unrepaired damage from a crash 18 months earlier.

In October, Hawaii recorded 106 traffic fatalities in 2025, more than all deaths in 2024. Although it is unclear how many of the deaths were related to road problems, Chen said the grim trend underscores the timeliness of the dashboard program.

Building a large AI database

The City of San Jose is reporting significant early success in identifying potholes and debris by simply installing cameras on a few street sweepers and parking enforcement vehicles.

But Mayor Matt Mahan, a Democrat who founded two technology startups before entering politics, said the effort would be more effective if cities contributed images to a shared AI database. Mahan said the system can recognize previous road problems, even if they occur elsewhere.

“You realize, ‘Oh, this is actually a cardboard box wedged between two parked vehicles, and this is considered debris on the road,'” Mahan said. “We could wait five years for that to happen here, or we could do it right away.”

San Jose officials helped establish the GovAI Coalition, which was formed in March 2024 to help governments share best practices and, ultimately, data. In addition to Colorado, other local governments in California, Minnesota, Oregon, Texas, and Washington are members.

Some solutions are easy

Not all AI approaches to improving road safety require cameras.

Massachusetts-based Cambridge Mobile Telematics has launched a system called StreetVision that uses cell phone data to identify risky driving behavior. The company is working with state transportation departments to identify where specific road conditions make it more dangerous.

Ryan McMahon, the company’s senior vice president of strategy and corporate development, was attending a conference in Washington, D.C., when he noticed his street vision software showing a large number of vehicles slamming on the brakes on a nearby road.

The reason was that the driver did not see the stop sign until the last moment because the bushes were blocking it.

“What we’re looking at is the accumulation of events,” McMahon said. “That gave me an infrastructure problem. The solution to the infrastructure problem was gardening shears.”

Texas officials have been using street vision and various other AI tools to address safety concerns. This approach was particularly useful when we recently scanned 250,000 lane miles (402,000 kilometers) to identify outdated road signs that were long overdue for replacement.

“If something was installed 10 or 15 years ago and there was a work order written on paper, God help you trying to find that in some numbers somewhere,” said Jim Markham, director of accident data for the Texas Department of Transportation. “Having AI that can inspect and sort through that is basically a power multiplier that allows you to see much faster, wider and further than you can drive things.”

Next up is self-driving cars

Experts in AI-based road safety technology say that what is being done now is largely just a stepping stone to a time when the majority of vehicles on the road will be driverless.

Pittman, CEO of Brinsea, who has worked on Hawaii’s dashcam program, predicts that within eight years, nearly every new car will have a camera, whether there’s a driver or not.

“How do we look at the roads today from the perspective of grandma in a Buick and from the perspective of Elon and his Tesla?” Pittman said. “This is a very important nuance for the Department of Transportation and city governments. They are currently building the infrastructure for both humans and autonomous drivers, and they need to start bridging that gap.”

© 2025 Associated Press. Unauthorized reproduction is prohibited. This material may not be published, broadcast, rewritten or redistributed without permission.