No abstract

No abstract

No abstract

No abstract

While current military systems are functionally capable of adaptively aiding human operators, the effectiveness of this capability depends on the availability of timely, reliable assessments of operator states to determine when and how to augment effectively. This paper describes a response to the technical challenges associated with establishing a foundation for reliable and effective adaptive aiding technologies. The central component of this approach is a realtime, model-based classifier and predictor of operator state on a continuous high resolution (0-100) scale. Using operator workload as a test case, our approach incorporates novel methods of integrating physiological, behavioral, and contextual factors for added precision and reliability. Preliminary research conducted in the Air Force Multi Attribute Task Battery (AF_MATB) illustrates the added value of contextual and behavioral data for physiological-derived workload estimates, as well as promising trends in the classification accuracy of our approach as the basis for employing adaptive aiding strategies.