Untangling operator monitoring approaches when designing intelligent adaptive systems for operational environments

Abstract

An Intelligent Adaptive System (IAS) is a synergy between an intelligent interface and adaptive automation technologies capable of context sensitive interaction with operators. A well-designed IAS should enable flexible task allocation between the operator and the machine. Research suggests that the integration of real-time operator state assessment (e.g., performance, psychophysiology) can create a true 'human-in-the-loop' system, thereby minimizing deleterious performance effects such as overlooking automation failures and slowly reorienting to tasks. However, these research approaches apply a variety of methodologies to determine sensors, metrics, and overall system design when applied to real world tasks. This paper seeks to untangle these issues such that a more comprehensive framework for systematically evaluating the utility of cognitive state detection methods is attainable. © 2014 Springer International Publishing Switzerland.