States turn to tech in push for safer roads
Moelyn Photos via Getty Images
By
Chris Teale
California wants to use artificial intelligence to work out where its road deaths are and how to prevent them. And New York has turned to the tech to try and prevent over-height vehicles hitting bridges.
LAS VEGAS — Federal figures released earlier this month found that just under 40,000 people died in traffic crashes last year on U.S. roads.
The U.S. Department of Transportation’s National Highway Traffic Safety Administration said in its early estimates that 39,345 people died in 2024, which it said was the lowest fatality rate since 2019. While NHTSA officials said the downward trend is positive, others said it should not obscure the fact that traffic fatalities are higher than they were a decade ago.
“There's no outrage,” Carlos Braceras, executive director of the Utah Department of Transportation, said during a panel discussion at the Google Cloud Next conference last week in Las Vegas. “We kind of take it, we hold it, because it's this couple here, a couple here another day, and we move past it. We don't think about it until it affects us personally. This safety urgency, that I wish we all had, isn't felt.”
Amid concerns about ongoing road deaths, some states are turning to technology and artificial intelligence to try to make their streets safer, whether it be through better protecting vulnerable road users or taking steps to prevent other incidents.
California, for example, loses between three and four people a day who are biking or walking on its roads. Tok Omishakin, the state’s transportation secretary, said the California State Transportation Agency is in the midst of procuring a model that can analyze where fatalities happen, at what time and any outside factors that contribute to that like poor lighting, a lack of a sidewalk or crosswalk, as well as a driver’s speed.
Omishakin said by analyzing that data, he hopes that the model can then suggest countermeasures that could mitigate those fatal crashes, especially in areas with high numbers of accidents.
“It’s a little bit of both being reactive and predictive: reacting to what's already happened on the ground where we have these conditions, and also being able to be predictive, and saying this may happen in the future if you continue to design or have conditions like this, even though it hasn't happened yet,” Omishakin said in an interview at Google Cloud Next. “This may happen to make those corrections ahead of time and not wait on those conditions again.”
The state also is experimenting with how AI can better synchronize its traffic signals to avoid congestion at its intersections, which also have the most carbon emissions as cars idle in traffic. Omishakin said it represents a big departure from previous generations’ decisions about transportation, which reflected that Americans “unintentionally fell in love with the car.” Now, he said, those investments must be more focused on human outcomes.
“The last 15 to 20 years, there's been a little bit of a renaissance,” Omishakin said. “I've been part of this lean and push towards making transportation be about people — not about infrastructure or specifically a vehicle — but transportation being a means to an end for people, because that's essentially what it is. It’s redirecting our focus, our investments, our policies, to say, how do we help people and goods move safely and efficiently? Not just, how do we build the biggest, fastest bridge that's going to move cars as fast as possible.”
Other states have looked to technology to solve some of their most vexing safety issues and moved fast to do so. One of the biggest problems that the New York State Thruway Authority faces is when trucks and other large vehicles are too tall to get under their bridges and therefore crash into them. That can cause massive tailbacks and hours of delays on the 570-mile toll highway, which connects the state’s major cities and so is the “economic backbone of the state,” said Josh Klemm, the Thruway Authority’s chief information officer.
In response, authority officials use computer vision in their traffic cameras to measure vehicles as they drive past, even at 80 miles per hour. If the system determines that a vehicle is over-height, it alerts the authorities, and they can then dispatch the New York State Police to pull that vehicle over.
Currently, the cameras triangulate a vehicle’s height using four signposts set up in a rectangle across the travel lanes. In time, and as the technology evolves, Klemm said they would like to produce messages for roadway signs alerting drivers that their vehicle is over-height or even notify them automatically through their in-car entertainment system or their cell phones. That is still a couple of years away, he said, but for now the cameras and the signposts are accurate on a vehicle’s height to within a couple of inches, and it has been cheap to implement.
“To be able to do it for tens of thousands of dollars is a no-brainer,” Klemm said in an interview at Google Cloud Next. “To save millions of dollars by spending 10 grand, it's a home run.”
While it may be tempting to think that transportation officials are turning to technology to solve their problems in lieu of traditional solutions, Omishakin said it is not that simple. He predicted that traffic engineers will not become “obsolete,” but that “their responsibilities may be a little bit different in the future.”
“If we deplete human intelligence, and we hand this over completely to AI, I think we're going to create way more problems for ourselves than we possibly even have today,” he added. Klemm said these efforts and others show what is still possible in government work.
“The public sector really can do great things,” he said. “At one point in time, society as a whole had faith in the government to deliver services, whether social services or programs, that would make the lives of Americans better. Over time, for one reason or another, that confidence in government has wavered. It's great to be in a position like this, where if you believe in what you want to do, and you have the grit and resolve to not give up because someone tells you it's not possible, you can deliver and execute at a scale that was previously untold and drive real change into the lives of millions of people.”