Assessing Comfort During Human-Robot Collaboration Using Virtual Reality Scenarios
摘要
Maintaining comfort during human-robot collaboration is critical for the efficient use of, and engagement with, emerging robotic technologies. In this paper, we describe our efforts towards designing and implementing a study to investigate the impact of robotic behaviors on human teammate comfort within manufacturing scenarios. We first developed immersive virtual reality (VR) human-robot collaboration scenarios in which we manipulated virtual robotic arm behavior parameters. These experimental manipulations included robot arm movement speed, human-robot proximity, and inconsistency of speed or proximity (i.e., predictability). We found that the lowest overall comfort ratings were reported for the very close proximity and very fast speed robot arm manipulations, particularly for the unpredictable conditions. At the end of the experiment, most participants noted that the very close proximity scenarios made them more uncomfortable than the fast speed conditions, suggesting the virtual robotic arm was considered as social entity with respect to proxemics. These findings and approach combined can help elucidate human perceptions of robotic behaviors for extension to real-world collaboration scenarios, with applications in manufacturing and beyond.