Marine and underwater robots - Perception, modeling, control

Marine robots (both underwater and on the water) are booming. At the dawn of the development of vehicle fleets, they have now acquired a certain industrial and scientific maturity. They are now widely used, mainly for off-shore oil applications, military applications (surveillance, mine clearance, etc.) and scientific applications (oceanography, climatology, etc.). In this article, we provide an overview of existing vehicles (USV, AUV, ROV, glider, sailing robot, bio-inspired robot, profiler, etc.) and a detailed description of the hardware associated with their operation, enabling localization, perception and communication. For each category of equipment, the operating principles are recalled and the essential technical data are summarized, to enable the reader to correctly select the right equipment for his or her application, taking into account the technical characteristics and conditions of use.

This technological state-of-the-art is accompanied in a second part by the scientific tools used to control marine vehicles: modeling and control. The model presented is the one defined by SNAME (Society of Naval Architects and Marine Engineers) and widely used by the community. Based on this model, we have selected three controls, whose operation and key points of adjustment are described below. These are the PID control, the adaptive non-linear feedback control and the sliding speed control. Examples illustrate their use, and their advantages and disadvantages are explained. Finally, the article concludes with an experimental comparison of the three controls presented, applied to the depth control of a mini ROV. This allows us to compare the controls and highlight their limitations.