Automated Aquaculture Operations With Vessel-Mounted Robotic Arm: An Experimental Feasibility Study

Abstract

Future aquaculture operations demand a higher degree of autonomy as new aquaculture locations are established in more exposed environments. In this study, we evaluate the feasibility of automating traditional fish farm operations through the utilization of a vessel-mounted robotic arm. We focus on a novel approach for automating the removal of deceased fish. Our work encompasses the technical design of a hose attachment and hook mechanism, simulations replicating realistic vessel and net pen motions, realtime prediction of fish cage collar positions using autoregressive models, and motion compensation control for the robotic arm. Scaled experiments indicate the feasibility of the proposed concept from a control perspective. This research contributes to the broader research challenge in robotics of managing interactions between a robotic arm mounted on a mobile platform and an environment that is also in motion.