The Human-Automation Behavioral Interaction Task (HABIT) analysis framework

Abstract

Human-machine automation systems involve integration among automated agents and goal-oriented human operators. Automation fundamentally changes the nature of the cognitive demands placed on the human operator, while simultaneously increasing the potential for out-of-the-loop performance problems to impact system efficacy. As a novel human factors method for assessing these human-automation interaction-derived, out-of-the-loop, performance challenges, we introduce the Human-Automation Behavioral Interaction Task (HABIT) analysis. The HABIT framework considers the drivers of system performance in terms of both cognitive activity and human behavior. As an example, we apply HABIT to a tedious intelligence, surveillance, and reconnaissance task. HABIT demonstrates its effectiveness in guiding the modeling and annotation of cognitive activities and successfully predicting actions resulting in errors. System designers can apply HABIT to assess human-automation interaction before a system is fielded to proactively address and mitigate out-of-the-loop performance problems.