Obstacle Avoidance for Autonomous Ground Vehicles in Outdoor Environments

The paper presents an obstacle avoidance algorithm to be used for autonomous ground vehicles applications. The proposed method improves some of the limitations of the recently developed null space based behavioral control. The technique divides the problem into tasks, which are associated to increasing priority. Activities with lower priority do not interfere with those having higher priority. The scenario is supposed known only partially, and the complete environment is reconstructed during the mission, with the aid of stereoscopic vision sensors. The validity of the method is currently verified via computer simulations.