A pilot study of fatigue and situation awareness during simulated small satellite operations

Abstract

Fatigue is a concern as it decreases performance, increases operational risk, and generates significant annual costs. Of particular concern with small satellites is the operation of multiple robotic assets. SPHERES, a small satellite testbed, was used to examine the effects of fatigue and presence of an augmented display with situation awareness aids. All procedures were approved by the MIT Committee on the use of Humans as Experimental Subjects. Subjects (n=8) were trained on the system, then randomized to a sleep protocol order, which included 3 days of rest and 3 days of sleep deprivation. Subjects controlled two satellites in a space debris avoidance scenario. Dependent measures included survival time, strategy used, fuel usage, secondary task reaction time, and responses to a situation awareness questionnaire. An ANOVA found significant interaction effects of sleep protocol order and fatigue state on survival time (p=0.008). For these pilot data, a post-hoc test showed better performance on day two if order was rested on the first day and partially sleep deprived on the second day (p <0.05). A successful strategy to increase survival time in the space debris avoidance task was to cover a large area. An ANOVA showed an interaction effect of order and fatigue condition (p = 0.001) on the use of a correct strategy. For these pilot data, a post-hoc test showed the highest area covered on test day two if order was rested on the first day and partially sleep deprived on the second day (p <0.05). A chi-square test for independence found subjects that were partially sleep deprived on the first day received no benefit from the augmentation while subjects that were rested on the first day did receive benefit (p = 0.018). Augmentation had no significant effect on survival time. The study results suggest that a mental model was easier to develop if the subjects were rested on day one due to a better performance and use of a sound strategy on test day 2. Here the augmented display may have assisted mental model development on subjects who were rested on day 1, as it led to improved situational awareness of the satellite's system states; this ability to use the augmentation is associated with better test day 2 performance. These pilot subjects highlight that training guidelines for operating multiple robotic assets should permit appropriate rest during and after training to assist in mental model development and later superior performance.