Autonomous Aquaculture: Implementation of an autonomous mission control system for unmanned underwater vehicle operations
Henning Odeby Karlsen, Herman B. Amundsen, Walter Caharija, Martin Ludvigsen
Abstract
This paper proposes a mission control system for a remotely operated vehicle (ROV) such that it is capable of performing an autonomous underwater inspection operation within aquaculture. The proposed autonomous mission control system combines the classical Guidance, Navigation, and Control architecture (GNC) with a hybrid mission control architecture. The logic flow within the autonomous mission control system is visualized via a finite state machine (FSM). The proposed mission control system was implemented on an Argus Mini ROV and verified in simulations using the FhSim framework. Furthermore, the autonomous mission control system was successfully tested at SINTEF ACE, a full-scale industrial fish farm laboratory located off the coast of mid-Norway. During the experiments, the ROV performed an autonomous inspection of a net pen using a Doppler velocity log (DVL). The autonomous mission control system was able to handle loss of sensor measurements through reactive behavior. Furthermore, using deliberative behavior, the mission control system attempted to keep track of and manage the autonomous net inspection task. During the trials the system suffered from inaccurate estimates of the net-relative position of the ROV.