Abstract:
BUG algorithms are effective strategies for local path planning in unknown environments. This article presents a practical implementation of the InsertBug algorithm using the Robot Operating System (ROS) and highlights its challenges. The algorithm relies on laser sensor and odometry data to construct a locally optimal path in an unknown terrain. Its evaluation was performed in the Gazebo 3D virtual environment, employing the TurtleBot 3 Burger robot. The evaluation spanned three types of environments: mazes, settings with simple convex and concave obstacles, and office spaces. The algorithm was assessed based on the robot's overall traveled distance and accumulated turns in yaw rotations measured in radians. The findings demonstrate the effectiveness of the algorithm in diverse layouts. The implementation serves as a valuable resource to further advance autonomous navigation systems.