EmotAI · EmotAI: Emotion Driver State Monitoring System based on AI

According to the European Union (EU) statistics, in Europe alone road crashes kills around 500 persons every week, more than 25000 people die every year, and more than 135000 are seriously injured, with a monetary cost of around 280 billion EUR. To halt this uncovered pandemic, the EU has declared its ambitious long-term goal call ‘Vision Zero’ to reduce the number of casualties and accidents to zero by the year 2050. To advance towards this goal, a General Safety Regulation (GSR), has been recently mandated to install driver assisting devices by 2030 focusing on the major factors causing crashes: driver distraction, fatigue and drowsiness. To comply with these regulations automobile companies are eager to install advanced driver assistance technologies in vehicles for driver assistance and alerting. Current solutions however are limited to some specific aspects (e.g., sleeping, phoning, looking off,), without having a holistic understanding of the whole situation of the driver, and with a big rate of false alarms that generate more distraction and annoyance to the driver, that ultimately prefers to switch it off to avoid having continuous false alarms without any benefit. To tackle this important challenge, in this fellowship, we propose to develop EmotAI (Emotion Driver State Monitoring System based on AI) a non-invasive tool that aims to address driver state monitoring and alertness challenges, to maximize driver attention and safety. We will be doing this by integrating novel AI-based technology using deep neural networks with non-invasive sensors and cameras with the novel theories in cognitive psychology in driving that aim to combine the information from 4 main pillars: the attention level, the activation level, the performance level, and the external factors. Our unique innovative solution aims to provide for the first time a comprehensive, holistic understanding of the driver’s behavior and emotional status to provide timely alerts to prevent accidents.