ROS2-Based AMR System for Mapping and Navigation in Unknown Indoor Environments

The rapid advancement of autonomous systems and automation technologies continues to revolutionize industrial processes, aligning with the goals of Industry 4.0. This thesis presents an enhanced Autonomous Mobile Robot (AMR) system intended for advanced indoor navigation and exploration, building on the groundwork established by the FIXIT project at CIM4.0. The primary objective of this research is to develop and implement a robust SLAM (Simultaneous Localization and Mapping) algorithm utilizing the latest capabilities of ROS2 (Robot Operating System 2). A key focus of this study is a comprehensive comparison of different SLAM approaches using the Nav2 library within the ROS2 framework. This analysis covers various algo??rithms available in Nav2, including grid-based and topological mapping methods, as well as different localization techniques such as AMCL (Adaptive Monte Carlo Localization) and EKF (Extended Kalman Filter). The comparison evaluates these approaches based on mapping accuracy, computational efficiency, and adaptability to dynamic environ??ments. Based on this analysis, an advanced SLAM methodology is developed, integrating the most effective elements from the compared approaches. This custom solution leverages Nav2’s modular architecture and ROS2’s improved distributed computing capabilities, allowing for efficient path planning and map optimization. The entire system is imple??mented using ROS2, taking advantage of its enhanced tools for simulation, visualization, and real-world deployment. Rigorous testing is conducted in various simulated environ??ments using updated versions of RViz and Gazebo, which are now more tightly integrated with the middleware. These simulations demonstrates the robot’s improved capabilities in active exploration, obstacle avoidance, and efficient mapping, showcasing the benefits of this approach. Finally, in the carefully controlled laboratory environment at CIM 4.0, real-world experiments were carried out to assess the robustness and performance of the created AMR system. The outcomes show how accurately and dependably the AMR can navigate through a variety of situations on its own, including unknowns areas and dynamic barriers.