Motion Planning for Omni-Directional Vehicles

Intelligent, maneuverable, autonomous vehicles will play a critical role in future Army systems.Key to the success of these systems will be their ability to quickly reach goal locations by traversing through unknown, unstructured or even de-structured environments. This proposal addresses how Omni-directional capability can be best exploited throughcustomized planning and control. A hierarchical planner is proposed, where a global planner computes near optimal abstract paths based on what is known, with the abstract paths being refined to actual vehicle configuration sequences during execution. Theprincipal advantage of this approach is that the commitment to the actual path is delayed until the vehicle is executing and sensing in its local terrain. In this way, the vehicle adapts to unknown and unexpected events in the environment without the needfor global re-planning. Three objectives will be achieved during the proposed work: (a) A UML design of whole planning system will be produced, (b) a prototype local planner will be implemented in software, and (c) the local planner will be verified insimulation by comparing Omni-directional, Akermann and skid-steer vehicle configurations over some simple terrains. Effective customized planning software is the key to enabling the potential maneuverability of ODV to be realized.Autonomous ODV's have great potential for many applications within the Army, for planetary rovers and most important commercially, to provide enhanced capabilities for tasks currently performed by Ackermann and skid-steer vehicles, in construction,agriculture, service and transportation.