Human-Integrated System Simulation Analysis “in the Wild”

Abstract

Systems engineers often use model-based simulations to evaluate design concepts before building physical prototypes and running live experiments. However, their simulations typically lack infrastructure representing human behavior (e.g., cognitive tasks) and performance measures (e.g., workload). While human factors practitioners have developed simulation approaches that include such infrastructure, they typically employ languages, tools, and techniques that engineering teams cannot easily adopt given real-world budget and schedule constraints. To address this challenge, we propose a novel, lightweight approach to human-integrated system simulation analysis. Our approach enables the analyst to take a preexisting system model, add minimal human-model infrastructure, and run whole-system simulations that produce operator-workload and task-duration estimates, all using a common engineering language and tool, SysML and Cameo Systems Modeler, and any preexisting behavioral modeling technique. We demonstrate the approach using a simple, preexisting model of a fictitious space-telescope system.