Using Cognitive Workload Analysis To Predict And Mitigate Workload For Training Simulation

Constructive simulation-based training does not always go according to plan. Instructors observe scenarios and direct adaptive responses to unexpected events. Simulation operators typically use a graphical interface to monitor the scenario and generate specific interventions that implement the instructor's intent. These activities require the operators' attention and cognitive resources, particularly for complex directions and scenarios with multiple entities to monitor. We are developing automated tools to reduce operator workload, thereby increasing the cost effectiveness of training. We use cognitive workload analysis to identify situations in which operators would be overloaded, and to identify the factors that cause the overloading. By identifying workload "hot spots," we can target more effective automated support strategies.GOMS (Goals, Operators, Methods, Selection) is a theoretical framework for workload analysis that has been applied to many modeling and automation applications. We use GOMS to create a detailed characterization of a set of typical monitoring and intervention tasks for fixed-wing air combat training. This enabled us to synthesize predictive models for combinations of these tasks. The models indicate where operators are likely to have difficulty implementing instructor's intent, as well as where operators are unlikely to be able to complete their tasks correctly. We then used these predictive models to guide the development of tools for reducing operator workload, as well as experimental scenarios for evaluating the effectiveness of those tools.This paper discusses the approach to GOMS modeling, as well as the details of how we used the predictive GOMS models to identify cognitive workload limitations. Using illustrative examples representative of fixed-wing air combat training, we describe composition of workload models, construction of combined predictive models, and use of models for development and testing. (C) 2015 The Authors. Published by Elsevier B.V.