Currently, in order to increase both safety and performance of human-machine systems, researchers from various domains gather together to work towards the use of operators' mental state estimation in the systems control-loop. Mental state estimation is performed using neurophysiological data recorded, for instance, using electroencephalography (EEG). Features such as power spectral densities in specific frequency bands are extracted from these data and used as indices or metrics. Another

more »

... ting approach could be to identify the dynamic model of such features. Hence, this article discusses the potential use of tools derived from the linear algebra and control communities to perform an approximation of the neurophysiological features model that could be explored to monitor the engagement of an operator. The method provides a smooth interpolation of all the data points allowing to extract frequential features that reveal fluctuations in engagement with growing time-on-task.