A Data-Driven Action Plan for Safer Roads

Unlike traditional systems that react to historical trends, new analytics methods using physics and AI-based tools are able to analyze roadway behaviors such as phone handling, speeding, hard braking, and aggressive cornering. This allows officials to assess risk as it emerges, enabling them to spot dangerous conditions before they result in injuries or fatalities. Proactive instead of reactive decision-making marks a fundamental shift in road safety strategy.

These analytical sources are built on the same actuarially validated risk models that insurers have used and regulators have approved for decades to predict crash risk. Studies have confirmed that one key method, telematics risk assessment (TRA) is built on procedures that can accurately predict crashes. The method uses key risk metrics to help predict areas for further analysis. For example, drivers with the highest levels of phone distraction are 240% more likely to crash, while high rates of hard braking are associated with 103% higher expected losses, and excessive speeding with a 71% increase in predicted losses.

Road safety officials are already using predictive risk tools to uncover solvable problems that traditional crash data might miss, such as faded paint, poor signage or obstructed sightlines. These are rapid, data-informed interventions that reduce road risk within days.

Any data source used must be grounded in privacy, safeguarding individuals by aggregating and anonymizing patterns of risk. No personal or trip-level information should be used, and all data collection practices must preserve the rights and privacy of individual drivers.

A significant advantage of modern data analytics is the ability to rapidly evaluate an intervention’s effectiveness. This leads to continuous improvement of safety programs and builds effectiveness based on improvements in risk, allowing for quicker responses and resource allocation for effective interventions.