Effect of Robot Gazing Behavior on Human Comfort and Robot Predictability in Navigation

Autor: Margot M.E. Neggers, Peter A.M. Ruijten, Raymond H. Cuijpers, Wijnand A. IJsselsteijn
Přispěvatelé: Human Technology Interaction
Jazyk: angličtina
Rok vydání: 2022
Předmět:
Zdroj: 2022 IEEE International Conference on Advanced Robotics and Its Social Impacts (ARSO)
Popis: If robots that navigate around people are predictable, people can anticipate on their behavior. To do so, humanoid robots can employ gaze cues to indicate their heading, but this would typically interfere with other gaze cues like making eye contact with a human in the same environment. The current study investigated the effect of different gaze strategies on human comfort and robot predictability. Sixty-seven participants participated in an online experiment in which they watched 12 videos of a simulated robot with 4 gaze strategies (constant direct gaze, constant evasive gaze, direct gaze and gaze left, and direct gaze and gaze right) and 3 possible trajectories (left, right, straight). Results showed that participants preferred the robot to look in the same direction as its trajectory and evasive gaze was perceived as less comfortable compared to the other options. Based on these findings, we conclude that robots should first make eye contact to make people feel comfortable before they look towards their destination to be more predictable.
Databáze: OpenAIRE