Benefits Assessment of Autonomous Emergency Braking Pedestrian Systems Based on Real World Accidents Reconstruction

Despite the success in reducing Spanish traffic fatalities by 65 percent in the past decade (2004 - 2013), pedestrian fatalities only have diminished by 45% (decreased by 35% in urban areas). This paper describes the main findings of a coordinated study performed by INSIA-UPM aimed to assess the potential influence of two active safety systems, a brake assist system (BAS) and an autonomous emergency braking system (AEB), in vehicle-pedestrian collisions through reconstruction of real-world accidents occurred in the city of Madrid (Spain). A total number of 50 vehicle-pedestrian collisions have been in-depth investigated following a common methodology, including on the spot data collection, analysis and reconstruction to estimate the collision speed and the pedestrian kinematics. Every single case has been virtual simulated twice using PC-Crash® software: the first is a reconstruction of the real accident and the second is a simulation in which the operation of active safety systems is emulated. The performance of the BAS system acts together with the antilock braking system (ABS). The AEB system emulated in this paper through computer simulations is based on the DaimlerChrysler’s PROTECTOR system. The benefit is assessed in terms of both collision speed and Injury Severity Probability (ISP) by comparing the reduction of their values from the real conditions to the virtual simulations. The pedestrian ISP was estimated, depending on the collision speed and the pedestrian head impact point, using a specific application to calculate its value based on the results of head form impact laboratory tests. The findings show that in several cases the collision could be avoided by implementing the active safety systems (12% if the vehicle was fitted with BAS+ABS system; 42% with PROTECTOR system); and it would reduce their consequences in terms of the estimated ISP. It was also found that in few cases a low reduction of the collision speed would increase the head injury severity (10%). Further research should include injury information and/or estimation (HIC). Other limitations are the sample size (only one city and frontal collisions) and no unhurt accidents have been included. The injury severity assessment within this study only considers head impacts to the front surface of the vehicle, injuries provoked by subsequent impacts were not taken into account. Hence it can be an interesting subject for further research. This is new because: it is a prospective assessment of active safety systems and autonomous emergency braking systems; it is based on accurate reconstructions, highly detailed parameters; the behavior of the system is simulated according to design parameters. Multi-disciplinary approaches such as this study make the identification of critical parameters easier and simplify the development of practical solutions by quantifying their potential impact on future actions to improve pedestrian safety. The active safety braking pedestrian systems have a potential benefit in real conditions. It also has limitations so the authors cannot rely just on it. It has to act together with other passive features and the driver has to keep aware. This methodology can serve to test the benefit of forthcoming active safety technologies.