Autonomous Docking for Underwater Drones Using ArUco Marker Based Localization

Autonomous docking for underwater drones is crucial for various marine applications. This paper introduces an approach to autonomous underwater vehicle (AUV) docking using ArUco marker-based localization, as demonstrated by Project MIRA, at the Tau Autonomy Center, Norway. The method combines a control system with dual-feed computer vision based perception to achieve autonomy. The control system is based on a cascaded PID controller fine-tuned using various external sensors and optimization techniques. The vision system utilizes ArUco markers for 3D positioning. Our method uses visual markers to help the drone navigate to a custom docking pad. We achieved stable and reliable docking by combining advanced control systems using visual-inertial sensors. This work highlights the potential of using visual markers and smart control systems for underwater drone operations.