Definitions, Models and Measures of Human Workload

Abstract

In the study of man-machine systems, whether we deal with manual or supervisory control/decision situations, there exists a great need to model human behavior quantitatively, since in order to describe the overall behavior of a man-machine system, we must know the behavior of the individual sub-systems man, and machine. Most of the research is focussed on performance studies. Well known are the performance models based on the control theory, such as the cross-over model (McRuer (4)) and the optimal control model (Kleinman and Baron (3)), and on decision and detection theory (Green and Swets (2); Sheridan and Ferrell (5)). These models have been shown to be of great help in the design of man-machine systems. Relatively little has been done in the field of ``work load''. Some measurement techniques are available, and are used, for lack of something better, in practical situations, but the results thus far are rather poor. Since in the design of man-machine systems workload appears to play such an essential role, development of better analytical models and measurement techniques are certainly of high priority.