Automatic Control has been a subject of studies for the last twenty years. It involves many difficult problems that have to be solved: conflict detection, modelling of uncertainties on trajectories, clustering of 1-to-1 conflict to find unconnected n-aircraft problems, etc. . . Moreover, the n-aircraft conflict resolution problem is highly combinatorial and cannot be optimally solved using classical mathematical optimization techniques. The set of admissible solutions is made of many unconnected subsets enclosing different local optima, but the subset enclosing the optimum cannot be found a priori. In this paper, we present an automatic conflict solver and its implementation in an Air Traffic simulator, with statistical results on real traffic over France. This solver, which takes into account speed uncertainties and allows aircraft to fly on direct routes, solves every conflict on a loaded day, and gives each aircraft its requested flight level and departure time.