Abstract
Approximately 31% and 28% of all casualty accidents in Japan and the United States (US) respectively are rear-end collisions, encouraging OEMs as well as suppliers to introduce rear-end collision avoidance and mitigation systems. Current strategies vary from simple warning over brake support to fully automated braking. The efficiency of different assistance systems was recently evaluated (7). Based on German real world accidents (GIDAS, German In-Depth Accident Study) and taking into account the driver behavior, avoided collisions or reductions of the collision speed were estimated. The assessment of collision avoidance systems in Japan and in the US is of high interest. In the current work, we present a transfer of our results from the German market to Japanese and US accident situations, using data provided by JSAE (Society of Automotive Engineers of Japan) and data from the General Estimates System (GES, NHTSA) respectively (10). Accident avoidance / mitigation by rear end collision assistance systems were determined in a former work for more than 8000 accidents from the German GIDAS database. As data with the necessary in-depth information is not at our disposition for other countries, evaluation procedures could not be directly transferred. Generic accident types were therefore defined on reduced information content and GIDAS accidents were grouped according to these types. Inspired from the GES coding, 10 accident types describing a rear-end collision with a parked, stopped, slower or a decelerating vehicle were defined. More than fifty rear end crashes could be identified and were classified according to the generic accident types described above. The efficiency of different driver assistance systems in Japan and in the United States were then assessed supposing identical avoidance / mitigation for each of the generic accident type as in Germany. Accident avoidance for different assistance systems, taking estimated driver behavior into account, ranges from 12% (predictive collision warning) up to 20% (automatic emergency braking) of all accidents with casualties in Japan and 11% up to 20% of all accidents with casualties in the US respectively. The share of avoidable collisions within all rear-end collisions is similar to Germany, as the relative share of rear end accident situations (generic accident types) to each other is nearly similar for the US and also if left hand traffic is taken into account for Japan. In conclusion, a methodology was indicated to estimate system benefits based on analysis of in-depth data to less detailed datasets. The estimate was based on a production level system, applied to rear-end accidents. Overall collision avoidance/mitigation systems show important benefits in Japan and in the US.
Keywords: Predictive Emergency Braking System, rear end collision, accident avoidance potential, impact assessment, incident and accident analysis