A MCA motion-planner for mobile robots with generic shapes and kinematics on variable terrains

Abstract

In the present work, we describe a fast Motion-Planner for Mobile Robots. Considering robots moving with smoothed trajectories on variable terrains, we have developed an algorithm based on an anisotropic propagation of attracting potentials on a non-Euclidean manifold. The optimal collision-free trajectories are found following the minimum valley of a potential hypersurface embedded in a 4D space. The planner is very flexible: it can be use on a wide class of vehicles with different kinematics and with generic shapes. Because of the latest property, it is also applicable to plan the movements of generic objects (e.g. in assembly workstations in manufacturing industry) as in Piano Mover's problem. Thanks to the underlying Multilayered Cellular Automata (MCA) architecture, it is a distributed approach. This planner turn out to be very fast, allowing to react to the dynamics of the world, evolving toward new solutions every time the environment changes without to be re-initialized. © Springer-Verlag 2004.