Autonomous Docking System with Obstacle Avoidance for a Differential Drive Robot

Nowadays, autonomous mobile robots are widely used in many different fields, from industries to home environments. To ensure autonomy from human intervention, a navigation system that includes obstacle avoidance, as well as an autonomous recharging system capable of guaranteeing long term activity, is essential. This work is part of a project carried out by the InnoTech System company in collaboration with the California State University of Los Angeles. Their goal is to develop a fully autonomous differential drive robot capable of executing various tasks such as helping people at the San Diego airport terminal or delivering food to students in the University Citadel. The aim of this thesis is the development and the implementation of an autonomous docking system that allows the mobile robot to reach the charging station avoiding potential obstacles. The introductory chapters present a review of the research solutions adopted to solve the autonomous charging issue and a summary of some of the actual state of the art obstacle avoidance algorithms. Subsequently, the project is divided into four main parts: the detection of the station identified by an augmented-reality tag, the docking process, the obstacle detection exploiting point cloud and the obstacle avoidance function. Lastly, after a brief description of the mobile robot that has been built in order to evaluate the developed system, the work concludes by analyzing the results obtained during the test phases.