Evaluation of Automated Emergency braking systems for collision avoidance

From the past few years, there is a pressing need for implementation of automatic in-vehicle safety systems to avoid vehicle crashes and fatalities. Development of autonomous emergency braking systems (AEB) to detect and avoid collisions in such critical moments is of paramount importance. In this paper, AEB control systems are developed for a four-wheeler system that aims to detect pedestrians and/or vehicles and halts the ego vehicle based on the time to collision (TTC). This control system aims to react based on the real-time relative distance & speed of the ego vehicle to actuate appropriate braking force. Control systems developed in Altair ActivateTM are co-simulated with CARLA, a virtual reality simulator for autonomous driving research. Various scenarios including sudden pedestrian entry, low and high-speed car to car motion, urban high traffic density environment, changing weather condition etc. are simulated to study the robustness of the control system. Further, studies are conducted to evaluate the effectiveness of the systems by varying the field of view of sensor, activation threshold, distance etc. The results of this evaluation demonstrate the effectiveness of the AEB system in reducing the severity of collisions, and the system's ability to function correctly in different scenarios. Later, suitable modifications to the implemented control systems are identified to change the lane systems in case TTC is extremely short for sufficient braking application.