Evaluating Mental Workload Measures in Human-robot Collaborative Assembly

Abstract

This study assesses the efficacy of various cognitive workload metrics in human-robot collaborative assembly tasks using a systematic review and meta-analysis of literature from Scopus and Web of Science. Key metrics evaluated include physiological (EEG, GSR, HRV), subjective (NASA-TLX), and behavioral measures. Findings reveal that physiological measures, notably EEG and GSR (e.g., EEG with $\mathrm{p}\lt \mathrm{0. 0 1}$ and GSR with $\mathrm{p}\lt \mathrm{0. 0 1}$), are highly sensitive to changes in cognitive workload but are constrained by technical challenges. Subjective assessments, particularly NASA-TLX, provide valuable perceptual insights ($\mathrm{p}\lt0.05$), while behavioral metrics reflect task performance impacts. Integrating these metrics is essential for accurate cognitive workload assessments in industrial settings, enhancing both the understanding and management of cognitive demands.