Project Manager: Esten I Grøtli, SINTEF Digital
Partners: SINTEF Ocean, SINTEF Digital, NTNU IMT, Argus, Lerow, Kongsberg Maritime and AQS
Duration: Q2 2015 – Q4 2018
The aim of this project is to develop and demonstrate the autonomous functionality of a robotic underwater vehicle equipped with the tools required to carry out frequent operations on aquaculture facilities in exposed areas.
The frequent and full inspection of nets is a key task carried out at modern fish farm facilities, and will probably become increasingly important if this type of holding system is to be used in more exposed localities.
Inspections are carried out to monitor the integrity of net pens and prevent escapes. It is assumed that the use of improved technology will enable more effective and less expensive inspections than current manual approaches employing divers. The use of divers will also be more restricted at exposed localities due to weather conditions.
Net pens are flexible structures with variable geometries, and this places strict requirements on vehicle navigation and control systems in terms of completing comprehensive inspections and the prevention of collisions.
The use of machine vision for inspections is challenging due to variable light conditions, fouling and turbid water.
This project has been focusing on addressing some of the basic challenges faced by an autonomous inspection system.
- A laser-assisted camera system for the lightweight positioning of AUVs/ROVs
- Model-based and data-driven detection methods for holes in nets
- Net-relative positioning
- A conceptual study of the ROV-LARS system
- Use of an ROV equipped with autonomous functions for inspection, maintenance and repair in current net pen systems:
- Hole detection in nets
- Hardware/software to enable effective positioning relative to nets and other objects
- Net pen condition assessments