A Survey on the combined use of IoT and Edge AI to improve Driver Monitoring systems

As revealed by different reports, one of the most dangerous daily activities is driving. In fact, a driver must be totally focused on the road, and avoid all possible distractions that may cause disasters or tragedies. This danger is especially felt by professional drivers, like taxi or truck drivers, that drive every day for long distances. Although different initiatives have already led to the introduction of specific legislative rules to prevent accidents (e.g., speed limits, maximum driving hours, obligatory wearing of seat belts, etc.), unfortunately various car accidents are still registered daily on almost every road in the world. The Internet of Things is one of the paradigms that have been preventing the occurrence of these dangerous scenarios by introducing new sensors and devices that enable more accurate control over both road and drivers. Nonetheless, in recent years, the IoT domain itself is dealing with a new revolution fostered by the introduction of Artificial Intelligence and Edge Computing concepts. The present paper aims to identify, through a state-of-the-art review, the most interesting works that exploit the combination of Artificial Intelligence, Edge Computing and IoT techniques to improve the safety in the driving sector. After presenting the results of the analysis that confirm the interest of researchers in such a direction, the paper describes a first version of a prototype developed to demonstrate the benefits that the combination of AI and IoT can bring to society and propose some interesting future scenarios