Validating a Human Performance Model Without a Complete System

Abstract

Human performance modeling (HPM) has been effective for determining crew designs. Crew design includes determining the number of operators needed, the role of automation, and member task responsibilities required to operate a system. Crew models are built for target systems under development to better inform the design and the needs of each operator’s user interface. This is a Verification and Validation (V&V) challenge—it is difficult to obtain metrics to validate a model when the completed system does not yet exist for obtaining sample measures. This paper addresses the challenge with an implementation of the V&V approach applying it to an HPM that successfully received V&V Accreditation for a large United States Air Force (USAF) command and control system. This paper then compares the results of the validated model with NASA TLX workload metrics and discusses the V&V effectiveness. Finally, it recommends better methods for V&V for human performance models.