Advancement in Advanced Driving Assistance Systems tools: Integrating vehicle dynamics, environmental perception and drivers' behaviours to assess vigilance

Abstract

Most existing Advanced Driving Assistance Systems (ADAS) are monolithic and provide dedicated functionalities such as lane detection, speed limitation or eye blink detection. Furthermore, systems which assess drivers' physiological states are often built independently from systems that monitor car dynamics or perceive the environment. The integration of a wide range of functionalities and information related to the driver's behaviour, environment and vehicle dynamics could potentially improve the reliability of ADAS. This paper describes a novel ADAS that integrates a variety of existing functionalities to detect drivers' vigilance impairment. Our approach merges existing functionalities and data in order to build a better assessment of the current situation and then predict future situations. Our goal is to comprehensively integrate information about car dynamics, drivers' behaviour and environmental perception in order to increase the accuracy of hypovigilance prediction.