Analisi Comparativa di Algoritmi di Pianificazione del Moto per il Task Follow Me.

The automation industry applied to logistics is undergoing a profound transformation, driven by evolving technologies and market demands. Traditionally, intralogistics solutions relied on Automated Guided Vehicles (AGVs), ideal for structured environments like the food, beverage, and tissue sectors. However, the rise of eCommerce has created a need for more flexible and adaptable solutions that operate autonomously in unstructured spaces shared with humans and other machines. To address this shift, E80 Group S.p.A., a leader in industrial automation, has launched R&D on Autonomous Mobile Robots (AMRs). This thesis, developed within the company, focuses on the navigation capabilities of Dingo-O, an omnidirectional AMR by Clearpath Robotics, using Nav2 (Navigation Stack 2) of ROS 2, a modular open-source framework for scalable robotics applications. The study analyzes and compares Nav2’s planning algorithms in the “follow me” task, where the robot must track a target, predict its motion, set an optimal goal position, plan a path, and navigate while avoiding obstacles. By evaluating performance, limitations, and strengths, this thesis aims to determine the best planning plugin for this application and explore the broader potential of ROS 2’s navigation stack in dynamic AMR environments.