LiDAR-based Dynamic Path Planning of a mobile robot adopting a costmap layer approach in ROS2

In the last two decades, the deep technological developments in the fields of artificial intelligence, computing platforms and sensors boosted the use of autonomous mobile robots in complex environments, such as factories, offices, hospitals, airports and, more recently, also in homes. One of the first problems that agents have to overcome for a successful navigation is the avoidance of dynamic obstacles. The LINKS Foundation in Turin carries out research in different domains, such as AI, IoT, Cyber-physical and autonomous systems. This thesis was born as one of the first building blocks of a LINKS project: build up an autonomous system of internal mail delivery across its offices, based on a fleet of mobile agents. At a first stage Turtlebot robots will be used for this purpose; each of which shall achieve the navigation goal relying on LIDAR data only. The objective of this thesis is to implement an efficient navigation within these constraints, exploiting ROS2, the recently released version of the widely established Robotic Operating System. At the same time, this thesis work aims at helping to solve the open points about dynamic obstacles avoidance in ROS2 Navigation Stack (Nav2), raised by the ROS2 community. Indeed, the latter does not embeds a strategy specific for dynamic obstacles handling during navigation. A research about the autonomous navigation problem and the ROS2 functionalities has been carried out. Then, it has been proposed and implemented a method to integrate dynamic obstacles handling within the Nav2 stack available in ROS2, exploiting the concept of costmap layers. The proposed approach aims at laying the base for a highly modular pipeline for navigation, allowing future adaptation to specific use cases in LINKS projects. Finally, simulations in virtual environments have been carried out using Webots simulator and Rviz, outlining improvements marings and critical points for the future work at LINKS.