Traffic safety assessment traditionally relies on police-reported crash data, which are often considered the “gold standard” because they are directly related to deaths, injuries, and property damage. However, relying on historical crash data for predictive modeling presents significant challenges, as such data is inherently a “lagging” indicator. Furthermore, accidents on main and local roads are statistically rare events, so it may take years to accumulate enough data to establish a valid safety profile for a specific road. road section. This sparseness paired with inconsistent reporting standards across regions complicates the development of robust risk prediction models. Proactive safety assessment requires “leading” measures: proxies for accident risk that are related to safety outcomes but occur more frequently than accidents.
In “From lagging to leading: validating hard braking events as high-density indicators of segment crash risk.“, we evaluate the efficacy of hard-braking events (HBEs) as a scalable surrogate for crash risk. HBEs are an instance where a vehicle’s forward deceleration exceeds a specific threshold (-3m/s²), which we interpret as an evasive maneuver. HBEs facilitate network-wide analysis because they are derived from connected vehicle data, such as time-to-collision. In contrast to proximity-based surrogates, which often require the use of fixed sensors, we established a statistically significant positive correlation between the rate of crashes (of any severity level) and HBE frequency. Virginia And California With anonymous, aggregated HBE information from Android Auto Platform.
