Autonomous Navigation with Obstacle Avoidance for a Car-Like Robot

Abstract

This work details the autonomous navigation with obstacle avoidance for a car-like robot. A control architecture is here proposed dealing with a low- and high-level controller, which are responsible for driving the actuators and for guiding the navigation, respectively. In order to minimize the car-like motion constrain effects, a forward/backward maneuver switched controller is proposed and here presented as the main contribution. Simulation and experimental results validate the proposal, i.e., the robot is capable to navigate in a semi-structured environment avoiding collisions in a reactive way.