Eye activity measures as indicators of drone operators’ workload and task completion strategies

We studied whether eye activity patterns in a simulated drone-operating task could be associated with workload levels and task completion strategies. Participants sent drones to suspected areas according to messages they received and according to self-initiated search. They were also required to validate whether suspected targets were indeed hostile prior to attacking them. We tested whether the number of suspected targets affected the number of eye transitions between task zones and whether it affected fixation duration in different task zones. We found that operators made fewer transitions between task zones as the number of targets increased. This was because they focused more on one zone and not on the others. Interestingly, the zone
they attended to relatively more was the one they needed for attacking targets and not the ones where targets usually appeared. This was probably because attacking required extended cognitive operations. Findings demonstrated that eye activity patterns can be used to infer task completion strategies and to identify workload levels, once these strategies are described. Workload levels and task completion strategies should therefore be studied by a combination of hypothesis driven and exploratory driven methods. Eye activity patterns can then be used as triggers for assisting overloaded operators.