Quantifying Autonomous Vehicle Pedestrian Interactions at Intersections
Autonomous vehicles (AVs) have the potential to enable a safe, efficient, equitable, healthy, and sustainable transportation system and communities. However, broad public adoption of AVs is predicated on the AVs’ ability to engage in safe and efficient interactions with other road users: conventional human-driven vehicles (HVs), pedestrians, and bicyclists, in our current infrastructure and traffic systems. While humans can anticipate and handle a range of other road users’ behaviors, unexpected behaviors that fall outside or in tail end of the range, can incite improper responses that could have ramifications for traffic safety and operations. Rapid advancement and adoption of AVs raises the important question of how well AVs can respect road users’ expectancies and coexist in harmony. Most studies on AV, while pioneering, involve SAE level 2 technology and are limited in scope, focusing mainly on AV (longitudinal) behavior through limited field experiments or AV-centric naturalistic driving. Thus, the state-of-the-art does not address AV-road user interactions, especially with pedestrians. This project will study how an AV interacts with pedestrians while making permissive turns at intersections.