Advanced Driver Assistance System (ADAS) Crash Safety Analysis via Onboard Monitoring System (OBMS) (Role: PI, Time: 2020–23, Federal)
Advanced driver assistance systems (ADAS) have great potential to prevent crashes and mitigate crash severity. With the goal, the Virginia Tech Transportation Institute (VTTI) team utilized data from onboard monitoring systems (OBMS) on commercial motor vehicles to develop realistic estimates of crash reductions (or increases) related to the adoption of ADAS. The objective of the project was to quantify the safety benefits of each type of ADAS represented using four years of historical, anonymized OBMS safety-critical event (SCE) data (crashes, hard braking, harsh lateral movement, tailgating, unnecessary hard acceleration, lane departure, and speeding). The primary focus of the analyses involves the crash rates of Class 8 trucks with Forward Collision Warning (FCW), Lane Departure Warning (LDW), Pedestrian Collision Warning (PCW), or Automatic Emergency Braking (AEB) compared to trucks without any ADAS. The Phase I datasets cover two consecutive years from July 2018 to June 2020, with at least 3,000 crashes and 3,000 near crashes per year. The analysis follows a retrospective cohort approach where each vehicle was grouped into the “with ADAS” or “without ADAS” cohort for a specific ADAS. The safety outcomes of each vehicle during the 2-year study period were compared to quantify the safety effectiveness of each ADAS. The study demonstrated that large-scale and objectively collected OBMS dataset can provides great opportunities to assess the benefits of ADAS for heavy trucks. Detailed information on near-crashes can be utilized to provide promising insights into the validity of near-crashes as a surrogate measure for future ADAS evaluation. This study, based on a rigorous cohort study design and novel crash and driving data, will contribute to the state of knowledge on the efficacy of ADAS for heavy trucks.