Active Participation in Highly Automated Systems: Turning the Wrong Stuff into the Right Stuff

摘要

A failure of human operators to take an active monitoring role in complex automated systems has resulted in operators who are less able to improve the efficiency and stability of a system and unable to make a transition from normal scanning behavior to the detection, diagnosis, and correction of system failures. Passive monitoring is common when operator training follows an associative or stimulus-response model. In this study, we manipulated operator-system participation and operator-operator communication to investigate the effects of increases in active participation on operator monitoring and problem-solving performance. 112 subjects worked as operators of a simulated process system. Operators worked in teams of two on both a monitoring task and, after the system failed, a diagnostic task. The results of this study suggest that active participation in the system improves both monitoring and diagnostic performance. In addition, active participation reduces boredom during monitoring, and stress while diagnosing a failure. Communication, on the other hand, was found to be a mixed blessing. Communication tended to facilitate performance of active participants, but degraded performance of passive participants. The implications of these results for system design, operator training, and future communication studies are discussed.