Session: 16-01-01: Government Agency Student Poster Competition

Paper Number: 149857

149857 - Safety Assessment of Adaptive Cruise Control Under Emergency Braking Scenarios 

Adaptive Cruise Control (ACC) equipped vehicles have witnessed significant growth in recent years due to their convenience, comfort, and safety benefits. ACC manages vehicle speed to maintain a safe following distance from the lead vehicle. Recent research has shown that ACC-equipped vehicles may be vulnerable to rear-end collisions (Mattas et al., 2020), especially during emergency braking scenarios, which raises concerns about the reliability of ACC in such situations.

In this study, we investigated the prospective safety performance of ACC-equipped vehicles, focusing on rear-end collision avoidance during emergency braking. We developed a metric called the Emergency Surrogate Safety Metric (ESSM) to measure collision-free safety criteria desired during emergency braking. ESSM calculates the difference between actual spacing and the minimum spacing (SS0) needed to avoid collisions. A positive ESSM indicates that the ACC vehicle can maintain a safe distance and avoid collisions during emergency braking.

This study uses data from empirical experiments conducted in Massachusetts (Li et al., 2021). The experiments consisted of a platoon of three vehicles driving in car-following mode. The experimental scenarios varied the ACC headways (minimum and maximum),  desired speed (low: 35 mph, medium: 45 mph, high: 65 mph), oscillation amplitude (small: 5 mph, large: 10 mph), low-speed cruise pattern (dip: accelerate immediately, long-cruise: cruise for 10-15 seconds at low speed), and deceleration and acceleration maneuver (mild, strong) to record the response of the ACC in different situation. The dataset consists of 96 experimental cases, with recorded data including position, speed, and acceleration values at 10 Hz using a high-accuracy GPS device (Ublox EVK-M8T) for one ACC model.

We applied the derivation of SS0 and ESSM to all leader-follower pairs of one ACC model from the empirical experiment. Specifically, for a leader-follower pair, each time instant in their trajectories presents an initial condition, and we assess “what if the leader suddenly applies an emergency braking from this time on”. If ESSM < 0, it indicates that collision will occur if the leader applies emergency braking from this instance onward. We selected the minimum value of ESSM for each trajectory pair.

The data analysis shows that the tested ACC system can exhibit unsafe behavior during emergency braking. In our analysis of 96 sample cases, 76 had a negative minimum ESSM value, indicating that 80% of the cases were at risk of a rear-end collision if the lead vehicle applied emergency braking. The ACC system is relatively safer in higher headway; however, it does not guarantee collision avoidance.  The ACC system does not maintain sufficient following distance from the lead vehicle for emergency braking scenarios. We further noted that the ACC controller's response time is crucial for vehicle safety, with higher response times increasing the risk of rear-end collisions.  The sensitivity analysis of the response time, emergency braking rate, and headway parameters ( time gap and space gap) suggest that an optimal range of parameters can be adopted for redesigning the ACC controller.

 As AVs are tested on public roads, ACC systems are expected to remain integral to these vehicles. Consequently, it is essential to redesign the spacing policy of ACC vehicles to be compatible with Autonomous Emergency Braking (AEB)  features to ensure safety.  The results of this study also necessitate further understanding of the interaction of different autonomous driving assistant systems to ensure vehicle safety is approached holistically, leading to more robust safety strategies for a wide range of driving scenarios and potential hazards.

Li, T., Chen, D., Zhou, H., Laval, J., & Xie, Y. (2021). Car-following behavior characteristics of adaptive cruise control vehicles based on empirical experiments. Transportation Research Part B: Methodological, 147, 67–91. https://doi.org/10.1016/j.trb.2021.03.003

Mattas, K., Makridis, M., Botzoris, G., Kriston, A., Minarini, F., Papadopoulos, B., Re, F., Rognelund, G., & Ciuffo, B. (2020). Fuzzy Surrogate Safety Metrics for real-time assessment of rear-end collision risk. A study based on empirical observations. Accident Analysis and Prevention, 148. https://doi.org/10.1016/j.aap.2020.105794

 

 

Presenting Author: Abhinav Sharma North Carolina State University

Presenting Author Biography: Abhinav Sharma is a PhD student in the Department of Civil, Construction, and Environmental Engineering at the North Carolina State University, Raleigh, NC, USA. He works to improve road safety and traffic flow through Intelligent Transportation Systems. His research includes automated and cooperative driving, transportation safety, and microscopic traffic modeling and simulation.

Authors:

Abhinav Sharma North Carolina State University
Danjue Chen North Carolina State University