Collaborative robots and user trust: The role of saccadic gaze, anthropomorphic motion, and repetitive training

The transition from Industry 4.0 to 5.0 increasingly emphasizes the importance of human-robot collaboration, where trust plays a critical role in interaction efficacy. This study analyzes the impact of specific robot design characteristics, saccadic gaze and anthropomorphic movement, on user trust in an industrial setting, as well as the effect of repetitive training. A total of 71 participants interacted with the Kinova Gen 3 robot across four distinct conditions: 1) with gaze and robotic movement, 2) without gaze and robotic movement, 3) with gaze and anthropomorphic movement, and 4) without gaze and anthropomorphic movement. The effect of repetitive training was examined through trust scores collected over these four conditions. Trust levels were assessed using the Human-Robot Interaction Trust Scale (HRITS) and analyzed through paired t-tests and repeated measures ANOVA, in accordance with the within-subjects design. The overall effect of repeated training did not reach statistical significance (F = 2.63; p = 0.058), whereas a significant linear trend was observed (F = 4.41; p = 0.039), suggesting a gradual increase in trust scores over time. Neither saccadic gaze (t = 0.06; p = 0.561) or anthropomorphic movement (t = 0.73; p = 0.466) showed relevant effects. These results suggest that, in industrial contexts, the familiarity provided by repetitive training may play a more relevant role in trust calibration than visual or kinematic anthropomorphic features.