DYNAMIC PATH PLANNING AND MAPPING USING SENSOR FUSION WITH ROS2

Abstract

This study presents a path planning and mapping system developed on the ROS2 platform, aimed at enabling mobile robots to navigate safely and adaptively in indoor environments with dynamic obstacles. The system uses only a LiDAR sensor for environmental perception and applies sensor fusion to improve localization using IMU and odometry data. The default DWB controller in Nav2 is utilized to perform local path planning, and costmap parameters are adjusted to dynamically update obstacle information. The performance of the system is evaluated through simulations in Gazebo, where costmap updates and obstacle types (static and moving) are compared across various test environments such as office, maze, and hexagonal layouts. The results show that dynamic obstacle detection and costmap updates significantly improve path efficiency and task completion time. The study demonstrates the potential of costmap tuning and sensor integration to enhance robot navigation in real-world scenarios.