Optimizing Human-Automation Team Workload through a Non-Invasive Detection System

In this project, Sonalysts will team with the University of North Dakota to establish the feasibility of a deployable, unobtrusive suite of sensors and data processing approaches collectively known as Cognitively-Oriented Physiological Indicators of Load, Operational performance, and Tension (CO-PILOT). CO-PILOT will provide real-time indicators of operator state that can be used to inform adaptive automation systems. The multi-dimensional nature of CO-PILOT will increase its predictive reliability and diagnostic accuracy. Its open system architecture will allow it to be dropped into a variety of systems and promote ease of integration. In Phase I, we will identify candidate measures for CO-PILOT and explore issues related to system architecture and evaluation. In Phase II, we will evaluate the candidate measures to identify the most efficient assessment set. With this set in hand, we will apply CO-PILOT to an authentic problem and illustrate its ability to enhance performance. In Phase III, we will integrate CO-PILOT with operational and/or training systems.