Nighttime safety of pedestrians: The role of pedestrian automatic emergency braking systems

The rise in vulnerable road user fatalities, e.g., pedestrians, is alarming. Nighttime poses unique challenges among the risk factors in pedestrian-involved crashes, accounting for approximately 74% of fatalities in the United States. Safe vehicles, a key element of the Safe System Approach, are a promising solution to improve pedestrian safety. Vehicle automation, particularly Pedestrian Automatic Emergency Braking (P-AEB) systems, can mitigate pedestrian-involved crashes. However, the P-AEB systems’ effectiveness, especially in darkness, has remained uncertain. Through analyzing the recent Insurance Institute for Highway Safety nighttime dataset from 2021 to 2023, integrated with vehicle data to create a unique database (1973 field tests), this study aims to understand the P-AEB systems’ effectiveness at night and explore correlates of their performance, especially the role of headlight technologies, employed sensors, vehicle propulsion type (Electric Vehicles (EVs) vs. non-EVs), and vehicle weight and size within the scope of controlled experimental data. Using a random-effects Heckman sample selection model with panel data, the study estimates the crash probability and impact speed in case of a crash, addressing the inherent panel structure of the data and unobserved heterogeneity among different vehicles and scenarios. Statistics indicate that P-AEBs stopped vehicles in 63.6% of the cases and, on average, reduced speed by 33.12% for the crashes. Results reveal that P-AEBs perform better in vehicles with light-emitting diode headlights than those with halogen headlights and may be relatively less effective in larger, heavier cars and EVs. Additionally, integrating camera and radar sensors improves P-AEBs’ reliability instead of relying solely on cameras. Future efforts can enhance nighttime pedestrian safety by focusing on advanced headlight technologies and sensor integration, improving P-AEBs’ prediction algorithms for various scenarios, and addressing systems’ limitations in larger, heavier vehicles and EVs.