OPTIMIZATION OF AUTONOMOUS MOBILE ROBOT NAVIGATION AND CONTROL BASED ON SLAM IN THE ROS ENVIRONMENT: AN EXPERIMENTAL ANALYSIS USING GAZEBO, RVIZ, AND THECONSTRUCT.AI SIMULATIONS

This thesis focuses on the optimization of autonomous mobile robot navigation and control based on Simultaneous Localization and Mapping (SLAM) in the Robot Operating System (ROS) environment. The research integrates state-of-the-art SLAM algorithms—EKF-SLAM, FastSLAM, ORB-SLAM, and GMapping—with classical and adaptive path planning methods such as Dijkstra, A*, D* Lite, and NMap. Experimental validation was conducted in Gazebo, RViz, and TheConstruct.ai simulation platforms, enabling analysis under both static and dynamic scenarios. The results demonstrate that ORB-SLAM provides the highest localization accuracy and mapping consistency, while NMap outperforms traditional planners in adaptability to dynamic environments. The findings suggest that a hybrid approach combining ORB-SLAM for localization with NMap for navigation offers optimal performance. The novelty of this work lies in the systematic integration of multiple SLAM and navigation algorithms within scalable simulation environments, contributing to more robust, efficient, and adaptable autonomous robot systems applicable in logistics, industrial automation, and intelligent transportation.