Empirical Study of the Braking Performance of Pedestrian Autonomous Emergency Braking (P-AEB)

<div class="section abstract"><div class="htmlview paragraph">Vehicle manufacturers are beginning to improve existing autonomous emergency braking (AEB) algorithms by adding pedestrian identification and avoidance capability. The Insurance Institute for Highway Safety (IIHS) has performed tests on eleven such vehicles; data are publicly available and were analyzed for this study. The first objective of this study was to compare Forward Collision Warning (FCW) engagement distance to target, pedestrian automatic emergency braking (P-AEB) brake application time, and incidences of impact across different manufacturers. It was observed that there exists a wide variation in FCW and AEB performance across manufacturers. FCW engagement distance tended to increase with test speed. Time from FCW engagement to AEB engagement was usually less than one second, with some manufacturer-specific variation. Incidences of impact tended to increase with travel speed, although some vehicles tested maintained constant number of incidences of impacts at all speeds tested. FCW tended to engage when the vehicle was farther from the target dummy as test speed increased, although the extent of that variation was vehicle specific. The second objective was to analyze the brake application characteristics of P-AEB, and how it varies across the different vehicles tested. A previous study of IIHS vehicle-to-vehicle AEB test data concluded that the algorithms generally employ a strategy of phased braking, wherein an initial phase of gradually ramping braking force (Phase 1) is followed by a period of steady-state braking (Phase 2). In this paper, the extent to which each P-AEB algorithm exhibits this 2-phase behavior was analyzed, and any unusual behavior was explained. It was observed that Phase 1 peak values and Phase 2 steady-state magnitudes were not strongly correlated to test speed for most vehicles tested, although there were some exceptions. Many test vehicles did not use the full theoretical maximum braking effort in Phase 1 or in Phase 2.</div></div>