Eye-movement patterns reflecting mental fatigue while working in augmented reality

Integrating eye trackers into AR and VR headsets provides an opportunity to continuously monitor users’ mental fatigue without additional sensors or periodic tasks. Mental fatigue has known behavioral and physiological correlates, but its ocular indicators, especially in AR environments, remain understudied and inconsistent in literature. To address this gap, we conducted an experiment with 47 participants using Microsoft HoloLens 2 during an interactive physics lesson in AR. Physiological signals (heart rate, skin conductance, respiration, and temperature) were collected to compute a composite mental fatigue indicator using ICA-based signal analysis and validated by comparison with changes in cognitive tests scores. Eye movement data were also collected to detect gaze events (fixations and saccades) and to find correlations with the mental fatigue indicator. The results showed a significant negative correlation between fatigue and average fixation duration, along with a trend toward an increased number of fixations, suggesting a pattern of more frequent but less focused visual scanning.