ABSTRACT: One of the effective ways to reduce road traffic causalities is to introduce Autonomous Vehicles (AVs) which have already started to emerge on the roads around the world. However, vulnerable road users (VRUs) may feel unsafe when they share roadway infrastructure with AVs (e.g. crossing a road in the vicinity of AVs). Research on the perception attitudes and safety of VRUs with respect to their interactions with AVs is limited and therefore, this paper aims to address this knowledge. This study then conducts a survey questionnaire revolted around the receptivity of VRUs toward road crossing around AVs. The survey data was analyzed by using data mining techniques such as Decision Tree (DT) and Random Forest (RF). RF identified important variables affecting the receptivity of VRUs with respect to varying levels of AV autonomy. These variables explored general attitudes toward AVs, concerns regarding AVs, and perceptions regarding VRUs crossing roads near AVs. In addition, the DT illustrated a tree model that provides pieces of information showing the relationships between the variables. Policymakers and automotive industries could use this information to widen their scope of VRU's receptivity factors for the efficient and safe implementation of AVs on roads.
Tarek Hassan, Xuesong Wang, Xiaoyan Xu, Bowen Cai, David Hurwitz, Mohammed Quddus. Analyzing the Receptivity of Vulnerable Road Users When Crossing Roads with the Presence of Autonomous Vehicles. Transportation Research Board 99th Annual Meeting, Washington D.C., USA, 2020. 1.12-16.