Towards the Implementation of a MPC-based Planner on an Autonomous All-Terrain Vehicle

Planning and control for a wheeled mobile robot are challenging problems when poorly traversable terrains, including dynamic obstacles, are considered. To accomplish a mission, the control system should firstly guarantee the vehicle integrity, for example with respect to possible roll-over/tip-over phenomena. A fundamental contribution to achieve this goal, however, comes from the planner as well. In fact, computing a path that takes into account the terrain traversability, the kinematic and dynamic vehicle constraints, and the presence of dynamic obstacles, is a first and crucial step towards ensuring the vehicle integrity. The present paper addresses some of the aforementioned issues, describing the hardware/software architecture of the planning and control system of an autonomous All-Terrain Mobile Robot and the implementation of a real-time path planner.