Regresar

Autonomous Intelligent Navigation for Mobile Robots in Closed Environments

Abstract:

Providing a map is mandatory for Autonomous Mobile Robots to be able to complete localization and navigation tasks, known as SLAM. Several SLAM algorithms which provides different quality maps have been proposed before but still issues related to map quality can appear while for accurate navigation high mapping performance is desired, therefore to be used in areas regarding health care through delivery and indoor control. For that reason, although several SLAM methods are available, the one provided by Cartographer ROS has been chosen for being one of the most recent, updated ones and has been taken into test with respect to the map quality provided. To accomplish that objective, the implementation of a simulation and experimental environment have been constructed in order to contrast between both mapping, localization and navigation results by using Turtlebot3 and Arlo Parallax platforms including LiDar and encoder sensors, with which the map created by the simulation would be the most optimum map as possible. As a result by using an RPLiDar A1, an acceptable map from the experimental procedure related to the optimized one was acquired. With which could be concluded that Cartographer ROS algorithm is satisfactory to be used for intelligent autonomous navigation purposes by providing high fidelity and effective maps even while demanding affordable computational power.