Sonar Based Outdoor Vehicle Navigation and Collision Avoidance

Abstract

Detecting unexpected obstacles and avoiding collisions is an important task for any autonomous mobile system. This paper describes an approach uslng a sonar system that we implemented for the autonomous land vehicle Navlab. The general hardware configuration of the system is shown, followed by a description of how the system builds a local grid map of its environment. The information collected in the map can then be used for a variety of applications in vehicle navigation like collision avoidance, feature tracking and parking. A slmple algorithm was Implemented that can track a static feature such as a rail, wall or an array of parked cars and use this information to drive the vehicle. Methods for filtering the raw data and generating the steering commands are discussed and the implementation for collision avoidance and its integration with other vehicle systems is described.