AMR system for autonomous indoor navigation in unknown environments

With the introduction and advancement of technologies now essential to industrial processes, technological evolution has significantly advanced the field of automation. The evolution of autonomous systems has made it possible to improve human work, facilitating collaboration with it while providing a substitute for handling the most demanding tasks. To achieve the goals of Industry 4.0, CIM4.0 has developed the FIXIT project, which aims to provide interactive support to the human operator in an industrial or logistics setting. The objective of this thesis is to develop an Autonomous Mobile Robot (AMR) system capable of autonomous indoor navigation through an unexplored and unknown environment. In order to fulfill the predefined task, a sensoristics system suitable for the environment is deployed. Exploiting the information coming from the sensors, an Active SLAM algorithm is implemented to extend the functionalities of the classic SLAM method to plan paths toward unkown spaces while mapping the environment. In order to carry on an analisys between different SLAM algortihms, a comparison of state of the art of Passive and Active SLAM solutions is performed. The resulting methodology is the adoption of an Active SLAM, in particular, the Google Cartographer method, which is used as the primary SLAM module to create submaps and efficiently conducting frontier detection in the geometrically aligned submaps generated by graph optimization. The overall system is developed using ROS (Robot Operating System) and has been validated in simulation using tools as Rviz and Gazebo. The functionalities of the developed Active SLAM algorithm have been tested in different simulation scenarios to prove the robustness and the efficiency of the solution. At the end, during the experimental phase, the performances of the real robot are evaluated in the CIM4.0 laboratory. The rover shows a great ability in actively exploring the environment, passing through narrow spaces and avoiding obstacles, while locating and mapping the discovered area.