VRoad: gesture-based interaction between pedestrians and automated vehicles in virtual reality

PubDate: September 2019

Teams: University of Oldenburg，OFFIS，Technische Hochschule Ingolstadt，University of New Hampshire

Writers: Uwe Gruenefeld；Sebastian Weiß；Andreas Löcken；Isabella Virgilio；Andrew L. KunSusanne Boll

PDF: VRoad: gesture-based interaction between pedestrians and automated vehicles in virtual reality

Abstract

As a third party to both automated and non-automated vehicles, pedestrians are among the most vulnerable participants in traffic. Currently, there is no way for them to communicate their intentions to an automated vehicle (AV). In this work, we explore the interactions between pedestrians and AVs at unmarked crossings. We propose a virtual reality testbed, in which we conducted a pilot study to compare three conditions: crossing a street before a car that (1) does not give information, (2) displays its locomotion, or (3) displays its locomotion and reacts to pedestrians’ gestures. Our results show that gestures introduce a new point of failure, which can increase pedestrians’ insecurity. However, communicating the vehicle’s locomotion supports pedestrians, helping them to make safer decisions.