· Contributors · Organizations ·
Sensitivity of Eye-Tracking Measures to Variations in Mental Workload while Learning to Operate a Physically-Coupled Robot
Virtual Program Session
Human AI Robot Teaming (HART)
DescriptionCollaborative robots are becoming more usable, versatile, and operable in close proximity with humans in a variety of industrial settings. However, the mental workload and motor-skill learning associated with operating these devices needs to be understood. We investigated the sensitivity of eye-tracking measures to variations in mental workload, as 18 participants learnt to use a physically-coupled robot for an object-manipulation task, under two levels of difficulty. We found that stationary gaze entropy (SGE), pupil dilation (PD) and fixation count (FC) increased, whereas gaze transition entropy (GTE) decreased with task difficulty. Additionally, SGE and PD reduced over learning, and FC increased with learning in the easy condition. SGE was the most sensitive eye-tracking metric to task difficulty. FC may be sensitive to changes in visuomotor strategies in addition to workload. Our results support the potential of eye-tracking to provide measures of workload as well as visuomotor strategies in physical human-robot interaction.