Comparison of Air Defense Warfare Task Performance with and without an Automated Task Manager Using a GOMS Modeling Tool

摘要

Small teams of human performance models perform tasks in a simulated Air Defense Warfare (ADW) scenario to compare two designs for an advanced multi- modal watch station (MMWS). In the original design, the models select and perform tasks based on their own knowledge of the on-going air warfare situation as acquired through visual search of individual tactical situation (TACSIT) displays and information windows and verbal communications among model operators. In a revised design, an intelligent agent Task Manager (TM) monitors the situation and generates a list of tasks on a separate display for the operator to monitor. While exercises on a prototype TM with human teams were confounded by user issues, the modeling study produced a quantitative comparison of task execution latency and total critical tasks between the two designs that showed generally improved performance for the revised design, thereby demonstrating the utility of human performance modeling for the evaluation of systems for complex team tasks.