Aquaculture field robotics: Applications, lessons learned and future prospects

TL;DR

This work surveys a decade of aquaculture robotics field experiments focused on enabling autonomous inspection, mapping, and interaction within dynamic net cages. By evaluating local and global localization methods, developing net-relative planning, and advancing robust control, the study demonstrates incremental gains in autonomy while highlighting persistent challenges from sensors, tethering, and fish behavior. The results show progress toward autonomous net inspection and safer obstacle avoidance, with adaptive control and safety frameworks reducing vulnerability to disturbances. The findings underscore the potential for precision fish farming to transform operations, while identifying critical gaps in SLAM, mapping of full net structures, and industry readiness for widespread deployment. The work also discusses future directions, including autonomous resident vehicles, docking solutions, and remote operation capabilities.

Abstract

Aquaculture is a big marine industry and contributes to securing global food demands. Underwater vehicles such as remotely operated vehicles (ROVs) are commonly used for inspection, maintenance, and intervention (IMR) tasks in fish farms. However, underwater vehicle operations in aquaculture face several unique and demanding challenges, such as navigation in dynamically changing environments with time-varying sealoads and poor hydroacoustic sensor capabilities, challenges yet to be properly addressed in research. This paper will present various endeavors to address these questions and improve the overall autonomy level in aquaculture robotics, with a focus on field experiments. We will also discuss lessons learned during field trials and potential future prospects in aquaculture robotics.

Figures (8)

• Figure 1: SINTEF ACE is a full-scale aquaculture laboratory consisting of several fish farms at various locations along the Norwegian coast. Pictured is the facility at Rataren, Frøya.
• Figure 2: Precision fish farming (PFF) aims at moving the operational principles of aquaculture from the current standard of manual operations and experience-based reasoning to a more control-oriented approach PrecisionFishFarming.
• Figure 3: ROVs are commonly used for net inspections in aquaculture. The image is from an experiment in SINTEF ACE where an ROV is performing autonomous net inspection using a DVL Amundsen2022.
• Figure 4: Mapping during autonomous net inspection using multibeam sonar Cardaillac2023.
• Figure 5: Online estimation of net cage deformation by using measurements from an ADCP and SBL transponders attached to the net Su2021.