Introduction
Biofouling is the growth of organisms such as algae, mussels and hydroids on submerged marine structures. In the sea-based aquaculture industry, biofouling is a challenge as it causes reduced oxygen flow through the fish cages, net deformation , and increased wear and tear on the nets and mooring systems
due to occlusion of the nets. Underwater robots utilizing pressure cleaners are often employed to remove the biofouling. This poses a risk to fish health, as harmful cleaning waste is released into the water. An alternative to pressure cleaning is to use brush-based grooming robots that gently disturb the net at regular intervals and hence prevent the biofouling from establishing
. Net inspection and monitoring of the environment is also of importance to the fish farmers. The project Netclean 24/7 (RCN 296392) has for the last 3 years develop ed the autonomous underwater robot
’Remora’ for net cleaning and monitoring. This work presents the results of the
Netclean 24/7 project and includes an analysis of the robots’ functions, modelling of the robot dynamics, autonomous control functions, and a
docking station design for permanently resident aquaculture robots.
Materials and methods
Before the development of the robot and its autonomous functions could begin , a thorough analysis of the requirements of the robot was performed. The SEATONOMOY method
was used for this purpose. This iterative method aims to define all the required operations the robot must be able to perform, then breaks these operations down into smaller sub-tasks . Once completed, the SEATONOMY method provides the designers with a comprehensive overview of the requirements for all software and hardware necessary to complete the robot
.
Developing autonomous functions for a robotic system is deemed easier if one has a mathematical model of the robots’ dynamics, and the environment it operates in. As s uch, a model was developed for the Remora robot moving on a dynamically changing fish cage
.
The autonomous functions of the robot were developed through a series of simulation experiments utilizing the mentioned model. The robot required algorithms that could ensure path following and obstacle avoidance. Based on the investigation of s everal such algorithms , the Elastic Band Method (EBM) was chosen. This method conceptualizes the robot’s path as a rubber band between a start point and an end point. If an obstacle appears, the path wraps around the obstacle, much like an elastic rubber band would do in the same situation. T o ensure that the robot maintains a safe distance to obstacles, a series of bubbles are placed along the path. The bubbles radii are chosen such that the inside of a bubble is considered a collision free area.
To realize fully autonomous and resident robotic systems in aquaculture a subsea docking station suited for aquaculture sites is required. An extensive study on existing subsea docking stations for autonomous underwater vehicles (AUVs), as well as a concept design for aquaculture use, were performed as part of the Netclean 24/7 project
.
Results
The results of the Netclean 24/7 project include an extensive analysis of the requirements and specifications for an autonomous, resident biofouling prevention robot. Further, a mathematical model
has been developed to describe the dynamics of crawling robots. Advanced motion planning and obstacle avoidance algorithms has been developed and extensively tested, both in simulations
, and laboratory experiments. A concept for a subsea docking station for aquaculture robots has been designed based on identified requirements from subsea docking stations used in other domains.
shows a simulation of the motion planning and obstacle avoidance utilizing the mathematical model. The model captures the interactions between
the robot, the net structure, and the ocean currents, while the motion planner is able to re-calculate the path to avoid obstacles.
shows the results from the laboratory trials. The robot was able to avoid both static and dynamic obstacles in a controlled environment .
shows the subsea docking station concept for residential aquaculture robots. The concept is based on wireless charging and data transfer.
Conclusion and future work
T he Netclean 24/7 project consists of project owner Remora Robotics and partners SINTEF Ocean, Nordlaks AS, Xylem Aanderaa AS, and NTNU.
The Remora robot is currently operational at several active fish farming locations in Norway and field trials implementing the results of the Netclean 24/7 project are planned in 2022.
References