Effectiveness of Advanced Driver Assistance Systems in Preventing System-Relevant Crashes 2021-01-0869

This retrospective cohort study uses survival analysis to estimate the effectiveness of Toyota ADAS in helping prevent system-relevant crashes. Toyota production data were merged with police reported crash files from eight U.S. states for crash years 2015 up to 2019 by 17-digit vehicle identification number (VIN). System-relevant crash scenarios included: striking vehicle in front-to-rear, single vehicle run-off-the-road, same-direction sideswipe, head-on, and pedestrian struck. The study vehicle cohort included 11 Toyota/Lexus models, model years 2015 through 2018, sold in the eight study states. ADAS technologies studied included automatic emergency braking (AEB), lane departure warning (LDW), lane keeping assistance (LKA), blind spot monitoring (BSM) and pedestrian automatic emergency braking (PedAEB). Among the study cohort of 2,394,913 vehicles, police reported 308,490 crashes. The crude crash rate ratio (CRR) was 0.61 for AEB-equipped versus non-equipped vehicles. However, the CRR does not adjust for differences in ADAS-equipped versus non- equipped vehicles. To adjust for group differences (confounding factors), Cox proportional-hazards (CPH) regression modeled the relative risk (hazard ratio, HR) of being in a system-relevant crash for vehicles with versus without the ADAS. CPH modeling found that AEB-equipped vehicles were 43% less likely (HR=0.57) to be the striking vehicle in a front-to-rear crash compared to non-equipped vehicles. The analysis was also stratified to look at the effect in intersection versus non-intersection crashes. BSM-equipped vehicles were 4% less likely (HR=0.96) to be involved in a same-direction sideswipe, though the differences were not significant (p=0.252). LKA-equipped vehicles were 9% less likely (HR=0.91) to run off the road. LDW and LKA did not have a significant effect on risk of same-direction sideswipe or head-on crash. PedAEB-equipped vehicles were less likely to hit a pedestrian (HR=0.84), but the hazard ratios were marginally significant (p=0.114). This study contributes new evidence of the effectiveness of ADAS in helping prevent system-targeted crashes.