“As far as Njord Aqua knows, we are the first to do such detections in real time,” says Njord Aqua Director Øyvind Lernes. “Fish escaping is one of the biggest challenges in the fish farming industry today, and this technology can contribute to significantly reducing the risk of this.”
In addition to being able to detect holes in real time, the software can boast of being independent of the type of equipment.
“The camera on the underwater robot (ROV) films the fish cage net and a machine learning algorithm analyses the images in real time on another computer. This means that if you have a video stream available, the method can be run, regardless of where it comes from,” says Research Manager Sveinung Johan Ohrem at SINTEF.
In this way, the time and resources needed for cage inspections can be greatly reduced, and it can make everyday life easier for those who work on the facilities:
“This depends on the size of the cages, but we assume that this can reduce the time spent by around 20%. Today, such inspections require the use of people and boats at the facilities. In the future, the task can be carried out by the drone alone,” Lernes explains.
Another advantage is that the inspections become more accurate. The software is already trained on thousands of images with varying visibility, light conditions, angles and fouling. And the more it is in use, the better it gets.
Good news for the industry
Njord Aqua sees several advantages with the software:
“The technology initially makes it possible to ensure that the operator is aware of all irregularities. It helps to remove a number of disadvantages in relation to human-to-machine problems, and it makes it possible to move on to more autonomous monitoring of underwater infrastructure.
“In particular, the technology will provide major advantages for facilities that are submerged in deeper water, and which therefore are not as easily accessible for diving inspections. In addition, the technology helps to a large extent to reduce costs, which is necessary for sustainable development.”
Ohrem at SINTEF adds: “Now that we have proven that it is possible to detect holes in real time from an ROV running a normal inspection round on a cage, we imagine that this could become the new norm for such inspections. In the ongoing project with Njord Aqua, Måsøval and NTNU, we are also working on implementing autonomous control methods for the ROV, which will make the inspections even more efficient.”
Transferable technology
Both Njord Aqua and SINTEF envision that the technology can be used for the inspection of other offshore installations as well, such as offshore wind, ships and oil & gas installations.
Win-win cooperation
The collaboration between Njord Aqua and SINTEF came about via an Innovation Project in Business (IPN) from the Research Council of Norway (RCN). The purpose of an IPN is to support companies that, together with partners, want to carry out a project that includes research-based innovation and/or that will demonstrate new technology.
“As a small company, we would not have had the opportunity to develop this without help and support from IPN and the Research Council of Norway. The scope, costs and risks are far too great for that. Furthermore, this is very advanced technology that requires a very high level of expertise to develop, and preferably from several professional fields. An entire team is therefore necessary to solve the task - something that SINTEF has contributed to,” Njord Aqua says.
“In IPN projects, SINTEF Ocean works closely with industrial partners to solve actual problems together. This is very valuable because we help to change processes and methods in the industry, and we contribute to developing technology that is rapidly taken into use. The close communication with the industry throughout is very valuable, since we then adopt the development and implementations to meet the customer’s needs,” Ohrem from SINTEF adds.
