Interaction between pedestrians and autonomous vehicles: Protocol analysis and preliminary results

Abstract

This research focuses on pedestrian crossing behavior when interacting with autonomous vehicles equipped with visual external Human Machine Interfaces (eHMIs). This study introduces an experimental procedure implemented in a CAVE-type pedestrian simulator in an approach that avoids the technical and ethical issues in doing such experiments in a real-world setting. Two preliminary experiments were conducted: the first included both autonomous and conventional vehicles on the road, and the second included only autonomous vehicles with visual eHMIs. Initial results show that the visual features incorporated in the autonomous vehicle helped participants decide to cross the road more quickly.