Classification of Driver Steering Intentions Using an Electroencephalogram

Abstract

For a functional driver assistance system to work property and provide cooperation between the driver and the vehicle, it must be configured to fit the preference of the driver. A brain—computer interface (BCI) provides communication between the driver and vehicle by translating human intentions, as reflected by brain signals represented in an electroencephalogram (EEG). This paper presents an algorithm for classifying a driver's steering intentions, based on a BCI that uses data from an EEG. Experiments were conducted with five able-bodied subjects, with varying driving experience, using a driving simulator. The off-line classification results show that the driver's steering intentions can be classified with an accuracy for about 65% for all subjects.