A matched case-control analysis of autonomous vs human-driven vehicle accidents

A matched case-control analysis of autonomous vs human-driven vehicle accidents

18 June 2024 | Mohamed Abdel-Aty & Shengxuan Ding
This study investigates the differences in accident occurrence between Autonomous Vehicles (AVs) and Human-Driven Vehicles (HDVs) using a matched case-control design. The analysis is based on 2100 Advanced Driving Systems (ADS) and Advanced Driver Assistance Systems (ADAS) data, as well as 35,113 HDV accident records. The findings suggest that ADS-equipped vehicles generally have a lower chance of occurring accidents compared to HDVs in most scenarios. However, ADS accidents are more frequent under dawn/dusk conditions and turning conditions, with odds ratios of 5.250 and 1.988, respectively. The study also highlights the importance of improving ADS safety in challenging conditions, such as enhancing sensors and algorithms. The research provides insights into the factors contributing to accidents involving AVs and HDVs, informing future development in autonomous technology and safety enhancements.This study investigates the differences in accident occurrence between Autonomous Vehicles (AVs) and Human-Driven Vehicles (HDVs) using a matched case-control design. The analysis is based on 2100 Advanced Driving Systems (ADS) and Advanced Driver Assistance Systems (ADAS) data, as well as 35,113 HDV accident records. The findings suggest that ADS-equipped vehicles generally have a lower chance of occurring accidents compared to HDVs in most scenarios. However, ADS accidents are more frequent under dawn/dusk conditions and turning conditions, with odds ratios of 5.250 and 1.988, respectively. The study also highlights the importance of improving ADS safety in challenging conditions, such as enhancing sensors and algorithms. The research provides insights into the factors contributing to accidents involving AVs and HDVs, informing future development in autonomous technology and safety enhancements.
Reach us at info@study.space