A fuzzy expert system for the early warning of accidents due to driver hypo-vigilance

In this paper a fuzzy expert system for the prediction of hypovigilance-related accidents is presented. The system uses physiological modalities in order to detect signs of extreme hypovigilance. An advantage of such a system is its extensibility regarding the physiological modalities and features that it can use as inputs. In that way, even though at present only eyelid-related features are exploited, in the future and for prototypes designed for professionals other physiological modalities, such as EEG can be easily integrated into the existing system in order to make it more robust and reliable.