Evaluation of a psychophysiologically controlled adaptive automation system, using performance on a tracking task

Three experiments were conducted to evaluate the performance of a psychophy siologically controlled adaptive automation system. Subjects were asked to perform a compensatory tracking task while their electroencephalogram (EEG) was recorded and an engagement index was derived from the EEG, using the a lpha, beta, and theta bandwidths: beta/(alpha + theta) and beta/theta. In E xperiment I, EEG was recorded from three different sites: frontal, parietal , and temporal. Although tracking performance did not differ as a function of site, the number of task mode allocations was greater under a negative f eedback contingency than under a positive feedback contingency. This effect was seen primarily from frontal sites. Experiments II and III evaluated th e adaptive automation system, using extended runs under positive and negati ve feedback with either a slope (Experiment II) or absolute (Experiment III ) criterion used to drive the system. Using either criterion, performance w as found to be significantly better under negative feedback. Future evaluat ion and use of psychophysiologically controlled adaptive automation systems are discussed.