Workspace Sharing with Proximity-Aware Robots: a Pilot on User Perspective

This paper presents a study on key Human Factors considered in a Human-Robot Interaction (HRI) manufacturing scenario. We investigate user-perceived trust in collaborative robots, targeting crucial aspects such as acceptance, interaction fluency, cognitive workload, and usability. The experimental study is focused on a car door inspection and assembly task, where a human operator and a cobot operate side by side within a small shared workspace. The second link of the robot platform is equipped with 30 distributed proximity sensors that map the surrounding environment and detect nearby obstacles. Two distinct control strategies are evaluated for generating collision avoidance motions. The first strategy, Sensor Mounting (SM), leverages the sensors' mounting locations as control inputs to generate reactive avoidance motions, as described in [1]. The second approach, Whole-Body (WB), utilizes any point within the robot's geometric model, enabling both sensorized and non-sensorized links to respond to unpredictable events, as detailed in [2]. 24 subjects were involved in the experimental trials, performing assembly actions alongside a UR10e robot. Without prior knowledge of the control strategies employed, participants completed an online survey to rate their overall experience in both robot operating conditions (SM and WB). Results suggested that the WB proximity-aware controller did not compromise the system's perceived usability, trustworthiness, or efficiency. No statistically significant differences were observed among key subjective metrics (p > 0.05). Acceptance, usefulness, and satisfaction scores remained consistently high across both conditions. Finally, qualitative insights suggested users' preference for the WB control strategy, often described as more adaptive and responsive.