A time optimal control problem for the computation of the fastest collision-free trajectory of a robot surrounded by obstacles is presented. The dynamics of the robot is governed by ordinary differential equations. To build the collision avoidance criterion, the robot and the obstacles are represented as finite unions of convex polyhedra. Each polyhedron is described by a system of linear inequalities. This description combined with linear programming arguments allows us to include the collision avoidance as linear state constraints in the optimal control problem. The resulting system is then solved by a sequential quadratic programming method. (© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)