In this paper, an agent based docking system is addressed. Functional AUVs (Autonomous Underwater Vehicles) have a small agent to dock with. When the docking device is detected, the AUV stops close to the device and sends the agent. It docks using acoustic and visual feed back. For efficient docking, the omni-directional docking device and robust detection method using vision is proposed. For demonstration of the efficiency of the system, experiments were carried out in both indoor and outdoor tanks.


Risky and complicate underwater tasks have been replaced by ROVs(Remotely Operated Vehicles) and they have improved the efficient of the tasks as well as accuracy. But they suffer from their heavy deck support and restrictions of movement from the tethers.

AUVs(Autonomous Underwater Vehicles) have excellent mobility with light deck support. But their intelligence and power source is not enough to undertake ROVs' complicate and heavy duty tasks. Mistakes of robots can bring out the critical accident in working tasks which touch objects. In order to prevent accidents, the reliable communication link for monitoring and emergency intervention should definitely be secured. The only feasible communication method, however, is an acoustic communication link, which have not enough the transmission speed and reliability for this purpose. Currently, a wired communication link is required during missions.

For missions requiring frequent or regular interval tasks, such as maintenance of underwater structures, docking devices for AUVs could be installed in near the working area. When AUVs dock to the device, it provides a wired communication link and power so they can be operated as ROVs. SWIMMER AUV (Evans, 2001; 2003) is an example. This kind of AUV/ROV hybrid vehicles can be an alternative to overcome the above difficulties or an interim solution to develop fully autonomous tasks by AUVs.

This content is only available via PDF.
You can access this article if you purchase or spend a download.