ABSTRACT: Autonomous emergency brake (AEB) is an active safety function for vehicles aiming to assist drivers in emergency conditions. The system reduces machine involvement by applying automatic and severe braking at low Time to Collision (TTC) values. Even though collision cannot be avoided, it is assumed that kinematic energy is reduced at impact considering lower speeds compared to an evasion-free collision. The current system provides uncomfortable deceleration rates in emergency conditions. Emergency braking also produces high strain for vulnerable road user who experiences a motorized vehicle approaching at an undesirable high-speed. The motivation of this paper is to evaluate an alternative configuration of AEB systems, specifically for Vehicle-Bicycle (V-B) encounters. The improved concept of an AEB system includes applying user-preferred thresholds that control the frequency of the system engagement and the strength of the response. A measure of comfort is added in this study to improve engagement of the system in V-B encounters. Distribution of braking rates and adopted TTC values observed from the Shanghai NDS can serve as the basis to select appropriate threshold values that trigger activation of the AEB system. More than 200 V-B encounters of different types are observed for characterization of braking rates and corresponding TTC values. Simulation is applied to evaluate the performance of the new system. Based on the results, the TTC of the improved AEB system under the comfort threshold is more conservative (greater values) and the deceleration rate is smaller leading to more user-alike behavior. This framework can be replicable to adapt AEB systems to regional conditions reflecting intrinsic human behaviour.
Yanli Bao, Xuesong Wang, Andrew Tarko, Cristhian Lizarazo Jimenez, Vamsi Krishna Bandaru. Vehicle-Bicycle Encounters Analysis to Improve Autonomous Emergency Braking Systems Based on Naturalistic Driving Study Data. Transportation Research Board 99th Annual Meeting, Washington D.C., USA, 2020. 1.12-16.