Automation, task difficulty, and aircrew performance.

The effects of an automated system on team processes and performance were assessed in a laboratory simulation. Two-person crews were required to fly a complex emergency-response scenario under conditions of low and high workload. These flights were completed with or without the aid of an autopilot. The results indicated that the autopilot was effective in reducing subjective workload. However, the automation was associated with improved performance on only 1 of 4 performance measures. Furthermore, it was observed that problem-solving performance was worse in the autopilot condition during the high-workload flights. Investigation of crew process data indicated that workload savings afforded by the autopilot might have been invested in more explicit coordination. The results are discussed in terms of their implications for military aviators' performance, system design, and team training.