Artificial intelligence technology is being deployed to address America’s annual toll of 40,000 roadway deaths through real-time monitoring and crash prediction systems.
Rekor Systems, founded in 2017, uses AI and vehicle data to identify dangerous road conditions before accidents occur. The company aims to move beyond reactive responses to crashes toward proactive prevention strategies.
“It really is one of the biggest crises in the U.S., to see 40,000 people a year dying on our roadways,” said Paul Zamsky, Vice President of Rekor Systems. The technology focuses on “detecting where crashes have happened without having to wait for a 911 call or identifying the areas that are riskier.”
The system analyses multiple data streams including weather conditions, driving behaviours such as swerving or sudden acceleration, construction zones, and special events. This comprehensive approach extends beyond traditional analysis of historical crash data to provide real-time risk assessment.
Vehicle data serves as the foundation for identifying hazardous roadways and potential solutions. Zamsky emphasised that all information is collected anonymously and in aggregated form, with cars functioning as “virtual sensors” to monitor road conditions.
“That enables us really to see what is happening on the road without having to have physical infrastructure, millions of dollars of hardware and cameras and everything put on the side of the road,” Zamsky explained.
The technology has been tested in pilot programmes across the United States for several years, though specific deployment locations and effectiveness metrics remain undisclosed. The approach represents a shift from expensive physical monitoring infrastructure toward data-driven analysis.
However, questions remain about the system’s practical effectiveness in preventing accidents. Whilst identifying risky conditions may help inform traffic management decisions, the technology’s ability to translate insights into actual crash prevention depends on rapid response capabilities and infrastructure modifications.
Privacy considerations also merit attention despite assurances about anonymous data collection. The extensive monitoring of vehicle behaviour patterns raises questions about data security and potential misuse, even when aggregated.
For traffic safety advocates, AI-powered monitoring represents a promising tool in addressing persistent roadway dangers. The annual death toll has remained stubbornly high despite decades of safety improvements, suggesting that traditional approaches may require technological enhancement.
The success of such systems will likely depend on integration with existing traffic management infrastructure and the ability to provide actionable intelligence that leads to concrete safety improvements rather than merely identifying problems after they occur.
