REAL-WORLD AUTONOMOUS INDOOR NAVIGATION USING SLAM AND MULTI-GOAL EXECUTION WITH TURTLEBOT3

This work presents a practical system for autonomous indoor navigation using a TurtleBot3 Burger robot in real-world environments. Such a system plays a crucial role in experimenting with autonomous navigation during the early stages of robotics education. We integrated SLAM with adaptive multi-goal planning and navigation using the ROS Navigation Stack under ROS Noetic on Ubuntu 20.04. The system incorporates GMapping-based SLAM, AMCL localization, and custom Python scripts to control flexible goal sequences. Testing was conducted entirely in a physical environment, with Gazebo used only during early experimental stages. The robot’s performance was evaluated across different goal counts and obstacle configurations using custom experimental metrics such as time taken, path accuracy, success rate, and recovery behavior frequency. The results validate the robustness of our approach in realistic settings and offer reproducible insights for the practical deployment of ROS-powered mobile robotics.