Section: Application Domains
Traffic Safety Control
Participant : Serge Haddad.
The Intelligent Transport Systems (ITS) community tries to deal with the numerous challenges that arise when designing secure and reliable software dedicated to automatic transport systems.
Several recent ITS projects aim at providing assistance to drivers and deal with partially automated motorways. The community investigated first a fully automated infrastructure and vehicles approach (as in the PATH project  ) in the 1990's. That approach was then abandoned in favor of a new line of research and development activities, more centered on safety strategies to ensure properties such as Collision Avoidance or Safety Margin for Assistance Vehicles  .
This vision relies on cooperative systems where “road operators, infrastructure, vehicles, their drivers and other road users will cooperate to deliver the most efficient, safe, secure and comfortable journeys”  . Implementing such a system then follows a peer-to-peer organization where each vehicle must fully cooperate in a time-constrained and safety-critical environment.
In that context, many projects are dealing with safety-oriented applications based on sensors, communication devices and protocols as well as distributed traffic management systems involving cooperation between the infrastructure and vehicles  ,  ,  . Thus, reliability, flexibility in the design as well as safety are primary issues. Such systems are even more complex to analyze than previous distributed systems. Consequently, there is a need for a specific methodology and tools to design and analyze them.
We will focus on an approach for the modeling and analysis of the collision avoidance problems in critical short sections of motorways with the aim to check whether a control strategy exists depending on the parameters (speed, safety distances, etc.). We intend to cope with the undecidability of such problems by appropriate discretizations and with high complexity of the obtained systems by using elaborated data structures based on decision diagrams.