Document Type


Publication Date





Current Unmanned Aerial Vehicle (UAV) operator task demands are highly variable and unbalanced across team members, resulting in sub-optimal operator utilization which leads to mishaps. This has driven the Department of Defense’s desire for more flexible team structures and task allocation tools. Unobtrusive and continuous measures of operator state are needed to effectively allocate tasking to operators and prevent errors. Twenty participants completed two twenty minute supervisory control sessions where task load was manipulated by varying event frequency (e.g., information requests) and eye tracking data was collected. Pupillometry data revealed increased mean and maximum pupil sizes with increased task load and larger pupil size standard deviation in participants who performed poorly, compared to those who performed well. These results suggest that increased pupil size is indicative of increased processing demands and could be predictive of task performance within a complex environment where performance measures can be challenging to obtain.