Traffic Safety Systems based on Vehicular Networks

The integration of telecommunication technologies into the automotive industry will change the way we travel today. The different vehicles on the road will be able to exchange information through Vehicular Ad Hoc Networks (VANETs), opening a new world of possibilities. This work focuses on improving traffic safety using this novel technology. The assessment of techniques for vehicular networks involves high cost and complexity, and hence simulation has become a widely used research tool. We developed a framework that uses mobility scenarios in real cities to simulate the communication between vehicles, allowing the development of efficient protocols for warning message dissemination in urban environments. These algorithms are part of a broader architecture, called e-NOTIFY, which is able to detect traffic accidents and inform the emergency services. To achieve a full automation of this process, Artificial Intelligence techniques are used to estimate the severity of the accident and determine the optimal set of resources for the rescue operation. The development and evaluation of a prototype proved the feasibility of the system, and how it could help to reduce road casualties.