Human Preferences in Task Allocation and Their Impact on Stress: An Experimental Study in Human–Robot Collaboration

Abstract

Human-centered task allocation is an important aspect of human-robot collaboration because allocation decisions can affect both system performance and the operator’s experience. This thesis investigates how accepted, rejected, and predefined task allocations influence subjective stress, perceived autonomy, task performance, and physiological activation. In a within-subject experiment with 30 participants, data were collected using questionnaires, ECG-derived heart-rate and heart-rate-variability measures, system log files, and post-experiment interviews.The results show that explicitly rejecting user preferences led to a more negative subjective experience, including higher perceived strain and lower satisfaction and trust. In contrast, accepted preferences increased perceived control. Physiological responses did not consistently correspond with subjective assessments. Overall, the findings indicate that preference handling mainly affects the subjective experience of collaboration and that task-allocation decisions should be communicated transparently.