In order to estimate the global effectiveness of Advanced Driver Assistance Systems, injury mitigation has to be taken into account. This is achieved through the use of injury risk curves that allow to convert speeds from accident scenario simulations into injury level probabilities. The present paper presents a new set of pedestrian injury risk curves (for fatal, severe and slight injuries), based on a French accident analysis database, and highlights the methodology used to build the curves and their confidence intervals.
The following analysis is based on car‐to‐pedestrian real‐world collisions obtained from police reports in France. The cases selected were those relevant for an Autonomous Emergency Braking pedestrian system, i.e., car to pedestrian frontal collision. Injury probabilities were modelled using a polytomous CLOGLOG regression model, depending on the squared impact speed. Confidence intervals were computed by bootstrap methods, based on a 20,000 samples simulation.
The model shows a sharp increase of the risk of fatal injury above 70 kph impact speed. At 73 kph impact speed, the risk of fatal injury reaches 50%. The maximal risk of severe injury is reached at 60 kph impact speed.
The use of these curves can provide a better estimation of the safety benefits as they take into account the potential gain for every injury severity level.