Wall following for autonomous robot navigation

Abstract

This paper presents the design of a novel approach to Wall-Following using a single distance sensor. Our concept takes into account distance measurements and odometry information from the robot. This information is combined using a Kalman filter, to estimate the relative position of the robot to the wall. A tracking controller is used to steer the robot with constant forward velocity along the wall while correcting distance errors. Simulations and tests with a real track drive robot are used to validate the approach. The comparison to a conventional two-sensor setup shows comparable precision.