Automated Vehicles (AVs) have the potential to prevent 95 percent of pedestrian injury crashes in the US, particularly when a driver violation or pedestrian visibility occurred more than one second before crossing [1]. Vehicles equipped with Advanced Driver Assistance Systems, such as front collision prevention/warning, lane departure prevention, emergency braking, and adaptive cruise control, are already accessible for purchase by consumers. These systems are believed to provide safety benefits due to their ability to reduce human errors in driving and minimize the likelihood of accidents. For example, Scanlon et al. [2] found that lane departure warning and lane departure prevention systems could prevent 28 to 32 percent of road departure crashes in the United States under current road infrastructure conditions. Cicchino [3] observed a 27 percent reduction in front-to-rear crash rates and a 20 percent decrease in front-to-rear injury crash rates with the use of forward collision warning systems. Furthermore, Cicchino [3] noted a 43 percent decrease in front-to-rear accident rates and a 45 percent reduction in front-to-rear injury crash rates with the implementation of low-speed autonomous emergency braking. It is estimated that over 400,000 injuries and nearly a million collisions could have been prevented in 2014 if forward collision warning with autonomous emergency braking had been installed in all vehicles nationwide [3].

Recent safety studies have focused on comparing Autonomous Driving Systems (ADS) safety with that of human drivers. For example, Kusano et al. [4] compared Waymo (an SAE Level 4 ADS) rider-only crash data to human drivers, and found a human crash rate 6.7 times higher compared to the ADS for crashes that caused injuries, and 2.2 percent higher compared to the ADS for policed-reported crashed vehicle rates. In 2024, a working group of industry, academic, and insurance experts developed the Retrospective Automated Vehicle Evaluation (RAVE) checklist, which sets out 15 recommendations to ensure the quality and validity, transparency, and accurate interpretation of retrospective ADS performance comparisons [5].