APPROACH:
Massachusetts Institute of Technology (MIT) completed the development of a virtual reality based Robotic Workstation Simulator (RWS), including simulations of representative fly-to/grapple, autosequence, and track-and-capture tasks, and associated performance metrics. Brigham and Women's Hospital (BWH) and MIT completed the protocol development. Subjects were screened at BWH then followed a normal sleep schedule (eight hours per night) for two to three weeks. During this time, they received up to six sessions of robotic training as long as they achieved criterion performance during each session. Over a 20 month period subjects were screened for robotics aptitude and medical factors, and followed a normal eight hour night for several weeks while training on a battery simulated space telerobotics tasks. They then underwent a week of six hours per night sleep restriction, and entered a time-cue-free laboratory environment, where they were tested over a 12 day period, before and after repeated 9 hour slam shifts in their sleep-work schedules. Sixteen subjects completed the study. Robotics tasks included ISS-like fly-to/grapple, autosequence monitoring, and track-and-capture manual control tasks. An imbedded secondary visual detection task was used to assess spare attention and thus mental workload. Performance on proxy cognitive and psychomotor tasks was evaluated at intervals. By the end of the 6 hour robotics session, subjects had been awake for 18 hours. Physiological measures (plasma and saliva melatonin and EEG), subjective sleepiness, and alertness measures were taken throughout. After each slam shift, subjects were allowed eight hours of recovery sleep on the pre-slam schedule. The first slam shift performance study was performed in 90 lux white light, and served as a control. In subsequent three day cycles, subjects were tested under three treatment conditions, whose order was randomized across subjects: a) 90 lux blue-enriched white light, b) 90 lux white light and hourly caffeine (0.3 mg/kg), and c) 90 lux blue-enriched light and caffeine.

RESULTS:
Investigators have suggested to their colleagues in the Robotics Branch of the NASA Astronaut Office that tracking the performance in early training lessons with a broader range of objective measures (e.g., time needed to complete lessons) will be necessary to improve the predictive capabilities of the models. Other tests of working memory and dynamic spatial ability should be evaluated for their correlation with other task performance or hand-controller categories.

Liu AM, Oman CM, Galvan R, and Natapoff R. Predicting space telerobotic operator training performance from human spatial ability assessment. Acta Astronautica. 2013. November; 92(1):38-47. [DOI]

Fatigue

Melatonin

Reaction time

Sleep

Spatial behavior

Alertness

Cognitive function

Cortisol, plasma

Eyelid opening speed

Fatigue

+ Click here for information of how this experiment is contributing to the HRP's path for risk reduction.