Investigating and Implementing ROS-based Autonomous Navigation for Mobile Robots using SLAM Algorithms

Abstract

Navigation is one of the most challenging tasks for a mobile robot. Mapping and navigation are becoming crucial in areas like exploration and automated vehicles, especially in places difficult for humans to access. To aid navigation, a Light Detection and Ranging (LiDAR) sensor is essential for creating maps of unknown environments. Our project focuses on mapping an unknown indoor space using Hector SLAM and GMapping techniques within the Robot Operating System (ROS). We used the Rviz tool in ROS to display the scanning results, and the maps created with LiDAR and ROS accurately match the real environment, making them ideal for indoor navigation. Additionally, we developed a navigation application where the robot first maps its current location. Then, we direct the robot to a new spot, and it moves there while updating and completing the map. This process repeats until the entire area is mapped. Our final results show that the robot can build maps and determine its location in real-time.