Potential Benefits of Autonomous Emergency Braking Based on In-depth Crash Reconstruction and Simulation

The objective of this study was to estimate the potential effectiveness of AEB systems using simulation of crashes drawn from Australian indepth crash data. 104 crashes that occurred within 100 km of Adelaide, South Australia, were used to assess the potential effect of AEB systems. The crashes had been investigated at the scene, re-constructed to determine collision speeds, and in this study they were analyzed using simulation to estimate how collision speeds and injury risks would have been modified by each of several AEB systems considered. Crash types considered were rear-end, pedestrian, head-on, right angle, right turn and a proportion of hit-fixed-object crashes. Other crash types were thought to be less responsive to the effects of AEB and were not considered. The variation in AEB systems were described using several parameters: the range of the forwardlooking zone, the angle or width of the forwardlooking zone, the processing time for the system to respond to the road user or object in its path (latency), the time-to-collision (TTC) at which the system would intervene, and the strength of the intervention (the level of braking). The AEB simulation used information from the trajectory of vehicles in the 104 crash reconstructions to estimate what difference each system would have made to the collision speed in each case and for each AEB system considered. Injury risk curves were used to estimate changes in fatal and injury crash risk in each case. The reductions in risk were weighted according to the rate of crash involvement of vehicles, based on the patterns of crashes in New South Wales for years 1999-2009. The overall reductions in risk produced by the various AEB systems were substantial. Systems were predicted to reduce fatal crashes by 20-25% and injury crashes by 25-35%. Note that these estimates rely on assumptions about universal operability and reliability of systems.