Driving monitoring is one of the main requirements of the upcoming automated driving systems because drivers must still be alert to take over the vehicle control in case of system malfunction. Therefore, this study aimed to design an accurate, predictive and robust driver drowsiness classification system.