Productive Inconvenience: Facilitating Posture Variability by Stimulating Robot-to-Human Handovers

Abstract

Collaborative robots that physically interact with humans in an ergonomic and safe manner are essential to the future of industry. A common task across many industrial applications is robot-to-human handover, in which the location of object exchange is vital in cultivating a seamless interaction. Most prior work on computing these exchange locations aims to adjust human posture towards a better ergonomic state during a single handover. This procedure typically involves the robot estimating the human’s biomechanical properties, e.g. center of mass and base of support, before determining an optimal handover location according to some ergonomics assessment scale. In a similar vein, we compare two methodologies for object handover, whereby the handover location is computed to either "assist" or "stimulate" the human receiver. Unlike existing approaches, we posit that improvements in human posture can be derived by stimulating the receiver’s movement dynamics to facilitate posture variability, rather than constrain or stabilize it. To compare methodologies, we conduct a within-subjects study where participants perform 78 object handovers with a collaborative robot architecture. Our ndings indicate an improvement in ergonomics scores for the "stimulating" approach, hinting at the importance of productive inconvenience in long-term robot-to-human handover.