The growing development and commercialisation of marine surface vehicles have driven significant advances in their design and integration. These vehicles, equipped with onboard computers, communication systems, and various sensors, offer flexibility for multiple applications. However, despite progress in motion control techniques, a considerable gap remains between validating algorithms in simulation and implementing them in real-world settings due to environmental uncertainties and inaccuracies in system modelling. This chapter addresses this issue by describing an experimental implementation of a path-following control strategy combined with an extended state observer. The control architecture is developed in Robot Operating System 2 (ROS 2) using open-source software and validated on the Yellowfish platform. Two cleaning tools designed for this vehicle are also introduced, expanding its practical applications. The experimental results confirm the robustness of the proposed strategy against external disturbances and validate the feasibility of transferring theoretical models into real-world implementations.