AI street cams catch bad drivers without spying on their faces

Speed signs and occasional police patrols clearly aren't fixing intersection accidents in the US. Now, automated AI cameras are stepping up to enforce traffic rules around the clock, aiming to turn stop signs from mere suggestions into hard stops.

Two startups, Stop for Kids and Obvio.ai, are deploying smart vision systems at local intersections to cut down on rolling stops, speeding, and failure to yield—the very infractions responsible for about half of all car crashes nationwide.

The automated eye on the corner

Stop for Kids was sparked by tragedy. In 2018, cofounder Kamran Barelli’s wife and three-year-old son were struck by a distracted driver and thrown nearly 18 meters. Speed display signs in their neighborhood just turned into a racing game for local teens, and extra police patrols were easily dodged via driver group texts.

Barelli pivoted his software team to build an AI camera that tracks vehicles at night using digital bounding boxes. The system skips faces and car interiors entirely, focusing only on license plates. Clean stops are deleted right away, while rolling stops trigger encrypted footage tied to DMV records for automated municipal ticketing.

In a 2022 pilot in Saddle Rock, New York, stop-sign compliance skyrocketed from a dismal 3% to 84% in 90 days, eventually reaching 94%. Around 60 Long Island towns are now exploring deployments, pending state legislative rules.

Solar setups and the human in the loop

California-based Obvio.ai is taking a slightly different route. Founded in 2023, its gear is solar-powered and connects over 5G, meaning cities can slap them onto utility poles without tearing up sidewalks.

Obvio's edge AI detects cars, bikes, and pedestrians in real time, ditching all benign video within hours. Crucially, Obvio keeps a human in the loop: local officials must manually review flagged footage before any citation goes out.

During a pilot across Maryland's Prince George's County last summer, violations were cut in half in weeks. In the city of Bowie, officials calmed public backlash over automated cash grabs by issuing warning letters instead of actual fines during testing.

Civil liberties advocates aren't fully sold, though. The New York Civil Liberties Union noted that vendors rarely back these numbers with independent research, and warned against mission creep—noting that a simple over-the-air software update could repurpose safety cameras for surveillance.

Why it matters

For residents and drivers, these systems promise safer neighborhood crosswalks without having an armed police officer on every block. Obvio’s warning-letter grace period and human-review policy feel like the right blueprint to build community trust before springing fines on unsuspecting drivers.

For municipal tech, this marks a shift toward lightweight edge infrastructure. Instead of ripping up streets to lay fiber and power for bulky intersection cameras, cities can now spin up solar-and-5G units in an afternoon at a fraction of the cost.

We think the real challenge will not be the computer vision accuracy, but regulatory guardrails. As civil rights groups rightly point out, once pole-mounted cameras with powerful edge AI are widespread, the temptation to expand their use beyond stop-sign enforcement is just a software push away.