Stochastic Mapping Analysis for Automated Guided Vehicles in ROS

Dynamic mapping has become one of the issues that demand attention in modern autonomous systems when navigating in an environment. This study investigates AGVs performance in such settings to demonstrate the behaviour of an autonomous system performing mapping in dynamic conditions. This work attempts to compare the two most used navigation techniques namely, the SLAM Cartographer and SLAM Toolbox to illustrate the performances of SLAM algorithms in dynamic and changing environments. By utilizing state of the art technologies like Robot Operating System (ROS), Gazebo simulations and Rviz using Turtlebot3 robots which is equipped with LiDAR sensors, the research aims to improve the mapping accuracy by considering the computational efficiency, and the system adaptability in the environment. The results shows that Cartographer has outperforms SLAM Toolbox on most metrics, and thus receiving significant insights to improve AGV's performance when performing its given tasks.