Mesoscopic Management of a Fleet of Cybercars at a Crossroads

Bibtex entry :

@mastersthesis { 2007mehani_cybercars_crossroads,
    abstract = { In the context of driverless automatic road vehicles, one interesting and non-trivial problem is that of the passing of a crossroads. At an intersection between two roads, the risk of collision is much higher due to the presence of other vehicles not going in the same direction, and potentially crossing each other's path. It is necessary to find safe algorithms allowing every vehicle to pass this sensitive point without colliding. Moreover, it is important, while doing so, to keep efficiency in mind, for the vehicles not to wait forever before entering and passing the intersection. Every vehicle must pass the crossroad as fast as the safe possibilities allow. This 6 months internship, done with the Imara team, tried to come up with an efficient algorithm for crossroads passing. A simulator has been written, giving the possibility to test various algorithms and efficiency comparisons have been made in order to determine which algorithm was the most promising, and where were the behaviors that could still be improved. Finally, an algorithm based on a reservation of critical points has seemed to be the most interesting. Details about its inner working and implementation are given, as well as preliminary results. },
    address = { Gothenburg, Sweden/Rocquencourt, France },
    author = { Mehani, Olivier },
    keywords = { 2007, chalmers, crossroads, cycab, driverless, imara, inria, lara, msc, mscthesis, oliviermehani, simulation },
    institution = { Chalmers Tekniska Högskola/Inria },
    title = { Mesoscopic Management of a Fleet of Cybercars at a Crossroads },
    url = { http://olivier.mehani.name/publications/2007mehani_cybercars_crossroads.pdf },
    year = { 2007 },
}