A Human-Centered Evaluation of Visualization Techniques for Teleoperated Assembly Tasks for Non-Expert Users

This study compares two visualization methods for teleoperated assembly tasks regarding the performance and workload of human operators. The two methods were direct view and two-dimensional video stream. In a between-subjects study design with 42 participants, we evaluated the operator performance of a teleoperated (dis)assembly task and compared the workload that the two visualization methods put on their operators. Performance was measured by task completion time, workload was subjectively measured by NASA-TLX, and objectively by a secondary task (n-back task). The results show that indirect visualization by a video stream leads to significantly lower performance. However, no significant differences were found in subjective or objective workload measurements. The results of the evaluations emphasize the importance of system evaluation in specific use cases and aid in the development of intuitive and efficient human-robot interfaces for teleoperated assembly tasks for non-expert users.