Human display monitoring and failure detection: Control theoretic models and experiments

Two newly developed control theoretic models for human display monitoring and decision making are presented that use the information structure of the well-known optimal control model of human response. Experimental paradigms with six dominant task variables (i.e. number of displayed processes, bandwidths, event probabilities, field of view, process couplings, and failure couplings), deduced from vehicle and process control situations, are the basis for extensive validation studies including eye-movement recordings. The broad coverage of the paradigms and the high degree of data/model correspondence provide the predictive potential for the analysis, design, and evaluation of man-machine systems. The relation of these models to existing prediction schemes is outlined.