Training-effectiveness of a mixed reality system for human–robot collaboration in industrial settings

Abstract

The integration of collaborative robots into manufacturing has significantly transformed industry dynamics by enhancing the effectiveness and adaptability of production processes. This study investigates the potential of mixed reality (MR) technology to revolutionize employee training through immersive and interactive environments that seamlessly merge digital and physical realities. We introduce an MR training system for human–robot collaborative assembly tasks and evaluate its effectiveness as a training tool. The evaluation applies the following metrics: User Experience, Interaction Effectiveness, Affective-Cognitive Response, Technology Readiness, and Task Completion Time. Sixty-three participants with varied technical backgrounds, experience levels, and job roles completed a collaborative assembly task using the MR system. Our analysis includes (1) the overall effectiveness of the MR system on an absolute scale, and (2) the difference in effectiveness given the participants’ backgrounds. For the latter, participants were split into two MR-Affinity groups. Our results suggest that the MR system can effectively train people to jointly work with robots. Secondly, no significant difference between the two MR-Affinity groups was found except for the Task Completion Time. Both results together indicate that MR-based training for human–robot collaboration is generally useful and applicable to users from varied backgrounds.