Toward Virtual Reality-based Evaluation of Robot Navigation among People

Fabien Grzeskowiak (1) , Marie Babel (1) , Julien Bruneau (1) , Julien Pettré (1)

1 RAINBOW - Sensor-based and interactive robotics

This paper explores the use of Virtual Reality (VR) to study human-robot interactions during navigation tasks by both immersing a user and a robot in a shared virtual spaces. VR combines the advantages of being safe (as robots and humans interacting by the means of VR but can physically be in remote places) and ecological (realistic environments are perceived by the robot and the human, and natural behaviors can be observed). Nevertheless, VR can introduce perceptual biases in the interaction and affect in some ways the observed behaviors, which can be problematic when used to acquire experimental data. In our case, not only human perception is concerned , but also the one of the robot which requires to be simulated to perceive the VR world. Thus, the contribution of this paper is twofold. It first provides a technical solution to perform human robot interactions in navigation tasks through VR: we describe how we combine motion tracking, VR devices, as well as robot sensors simulation algorithms to immerse together a human and a robot in a shared virtual space. We then assess a simple interaction task that we replicate in real and in virtual conditions to perform a first estimation of the importance of the biases introduced by the use of VR on both a Human and a robot. Our conclusions are in favor of using VR to study human-robot interactions, and we are developing directions for future work.

Mots clés: Motion Tracking; Virtual Reality; Robots simulation; Human Robot Interaction

Domaines: Robotique [cs.RO]

Fabien Grzeskowiak, Marie Babel, Julien Bruneau, Julien Pettré. Toward Virtual Reality-based Evaluation of Robot Navigation among People. VR 2020 - 27th IEEE Conference on Virtual Reality and 3D User Interfaces, Mar 2020, Atlanta, United States. ⟨hal-02473081⟩