Objective: The aim of this study is to provide experimental evidence for the problem of complacency when using the AFCS (Automatic Flight Control System) of KUH-1 and suggest ways to prevent it.
Background: With the introduction of the modern cockpit automation systems, the risk of complacency effect has increased. Many studies have been conducted on the complacency, but no experimental studies have investigated the complacency in the actual aviation domain using autopilot. In this study, the impact of trust and expertise on monitoring behavior and the impact of monitoring behavior on flight performance are investigated. Automation induced complacency is inferred as a consequence of monitoring behavior.
Method: In the KUH-1 simulator, 19 pilots flew in the traffic pattern. During the flight, monitoring behavior data were collected with an eye-tracker, and flight performance was measured by whether to be off course and the action time in case of AFCS failure.
Trust in AFCS was measured by a questionnaire.
Results: Trust in AFCS did not affect monitoring behavior in the entire sample. But trust induced different monitoring behavior depending on the expertise. Pilots who were off course focused longer on the secondary task at once. The action time in AFCS failure was significantly related to the proportion of total dwell time in the secondary task and the pilots' flight hours.
Conclusion: The complacency can occur to both experts and novices, and can put them in a dangerous situation. Having a high level of manual flight skills and expertise, calibrating trust to an appropriate level, putting aircraft control as a top priority, and maintaining situation awareness through sufficient monitoring and transition between tasks will be ways to prevent and respond to complacency.
Application: This result can be used as an initial empirical study on the interaction between the automation system and helicopter pilots, and will be a basic study to safely operate the cockpit automation system at army aviation of Korea.