摘要: |
This report describes an experiment evaluating the effects of supervisory control automation on attention allocation while operating a train. The study compared two levels of supervisory control (partial and full) to manual control, in terms of how it affects vigilance detection and situation awareness. Human performance was measured using a human-in-the-loop train simulator. To evaluate vigilance, participants were asked to detect two types of automation failures and react to obstructions on the track. Situation awareness was measured using the Situation Awareness Global Assessment Technique (SAGAT) in which the simulation was suspended at periodic intervals and the subjects answered questions about the system. These answers were compared to objective measures of system performance. Attention allocation varied with the method by which supervisory control was implemented. In particular, attention allocation for speed control, a critical piece of information, varied with the two methods. Partial supervisory control, as implemented in this experiment, resulted in a narrowing of attention. The primary focus was on speed information. By contrast, use of full supervisory control resulted in participants spreading their attention more broadly, but at the expense of speed information. |