A novel approach for evaluating the potential benefits of motorcycle autonomous emergency braking in real world crashes

Motorcycle autonomous emergency braking (MAEB) was recently identified as a promising safety solution, the applicability of which was estimated to be one third of all motorcycle crashes. Further evaluations are needed to clarify the potential benefits of MAEB in real world crashes. In this paper, a new method is presented. The method involves firstly identifying the types of crash scenarios where autonomous emergency braking is likely to be applicable. Crash types are classified using the Definition for Classifying Accidents (DCA) codes. Secondly, for each selected DCA a number of representative crashes are chosen as reference cases (baseline) from real world motorcycle crashes in NSW and SA. A distribution of slightly modified cases (variants) from the baseline cases is then generated. Adopting the Monte Carlo method, the initial position of the vehicles, the reaction times and the type of reactions are randomly altered in each variant within predefined ranges. The effects of MAEB, in terms of impact speed reduction and changes in impact configuration are then estimated using computer simulations across the distribution of DCA crash types. The estimated outcomes for each DCA can then be extrapolated to provide an overall estimation of the potential benefits of MAEB using mass crash data from SA and NSW.