Autonomous technology for aquaculture-related IMR operations

PhD-student: Bent Oddvar Arnesen Haugaløkken

Supervisors:

Professor Ingrid Schjølberg (NTNU)
Professor Ingrid B. Utne (NTNU)

Duration: Q3 2016 – Q3 2019

Defence of thesis: 13 January 2020.


This Ph.D. study focuses on ways of developing autonomous technology for operations such as inspections, maintenance and repair in the aquaculture sector. This will mainly involve the use of inexpensive underwater robots and other equipment that can deployed inside conventional net pens.

There are many challenges linked to aquaculture-related inspection, maintenance and repair operations both inside and outside conventional net pens. These include the development of holes in nets that can lead to large-scale fish escapes. One of the causes of net wear and tear is the development of holes during washing. Such operations have in common that they must be carried out at regular intervals and are thus prime candidates for greater levels of automation.

Underwater vehicles can carry out autonomous inspections. If the vehicle is equipped with a robot arm, it is also capable of carrying out small maintenance and repair operations.

This Ph.D. study will attempt to demonstrate how a small observation class ROV can be deployed to carry out straightforward maintenance tasks. Guidance mechanisms have been developed that enable the vehicle to maintain a stable position in the water, even in weak currents.

Further work will address the autonomous grasping of objects in the water column with the aid of a camera mounted on a robot arm gripper. Experiments are planned in which object recognition will be attempted using a variety of methods. For the most part, these methods will involve object detection (models trained using machine learning), bar codes, colour and edge detectors.