Reliable Motion Plannning for a Mobile Robot

Abstract

Autonomous mobile robots need to be equipped with appropriate planification and control navigation systems in order to obtain safe behaviours. This study aims at implementing on a two wheeled mobile robot a robust autonomous navigation planning algorithm, which guarantees a safe and reliable path. By making use of all facilities which the robot operating system (ROS) middleware and the open motion planning library (OMPL) can offer, we implement on a mobile robot platform an autonomous architecture adapted to our problem. The planning navigation system makes use of the widely-used Rapidly-exploring Random Tree (RRT) algorithm, while the control navigation level is based on the go-to-goal strategy. As a novelty, our previously proposed reliable and safe navigation planning algorithm based on RRT principles and solved in an interval analysis framework, e.g., BoxRRT, is tested on a mobile robot platform. Through experiments, the interest of using such a robust navigation planner on an autonomous navigation architecture with uncertain localization is presented.