Exploring relationships of human-automation interaction consequences on pilots: Uncovering subsystems

Objective: We attempted to understand the latent structure underlying the systems pilots use to operate in situations involving human-automation interaction (HAI). Background: HAI is an important characteristic of many modern work situations. Of course, the cognitive subsystems are not immediately apparent by observing a functioning system, but correlations between variables may reveal important relations. Method: The current report examined pilot judgments of 11 HAI dimensions (e.g., Workload, Task Management, Stress/Nervousness, Monitoring Automation, and Cross-Checking Automation) across 48 scenarios that required airline pilots to interact with automation on the flight deck. Results: We found three major clusters of the dimensions identifying subsystems on the flight deck: a workload subsystem, a management subsystem, and an awareness subsystem. Discussion: Relationships characterized by simple correlations cohered in ways that suggested underlying subsystems consistent with those that had previously been theorized. Application: Understanding the relationship among dimensions affecting HAI is an important aspect in determining how a new piece of automation designed to affect one dimension will affect other dimensions as well.