Foto: DTU Aqua

Live TV from the seabed for more sustainable fisheries

Tuesday 15 Dec 20

Contact

Ludvig Ahm Krag
Professor
DTU Aqua
+45 35 88 32 06

About Norway lobster fishing

Norway lobster fishing is one of the most important commercial fisheries in Denmark. Approximately 130 commercial fishermen are engaged in Norway lobster fishing, and the Norway lobster industry generates annual revenue of EUR 500 million at EU level. 

Prohibition on discarding fish

Relatively small-meshed nets are used in Norway lobster fishing. It is therefore difficult to prevent other species such as small cod from being caught in the net, and this is a problem for the fishermen.

Since the beginning of 2019, all catches of virtually all species must be landed. Before then, the fisherman could discard the small fish. Now they have to be brought ashore, and they are deducted from the fisherman’s quota.

If a Norway lobster fisherman catches too many small cod, this may block him from catching all the Norway lobsters for which he has a quota.

Therefore, the fishermen are very interested in avoiding catching the wrong species. With the new system, the fisherman can interrupt fishing along the way if he sees that the quantity of Norway lobster is too low relative to the quantity of bycatches. 

A new camera system that can be installed in bottom trawls provides commercial fishermen with a tool for monitoring their fishing in real time. They can now for the first time see whether they are catching what they want, while the cutter pulls the trawl along the seabed.

The wheel house of a cutter is filled with technology and screens. The fisherman knows precisely where and at what depth the cutter is fishing, but the catching process is blind, and experience, skill, and, not least, a good deal of luck can be essential to the effectiveness of the fishing.

Now, a new camera system can broadcast live video footage from a trawl on the seabed, so that the fishermen can assess whether they are catching something and whether it is the right species and sizes.

“This technology allows the fisherman to monitor his process for the first time. This can make fishing operations more intelligent. It can optimize both the economic and biological sustainability of fisheries.” Says one of the researchers behind the system, Professor Ludvig Ahm Krag, DTU Aqua.


Video footage from a trawl on the seabed.

From blind trawling to a clear view
The system has been developed for trawler fishing. During testing of the system, particular attention has been paid to seabed fishing of flatfish and Norway lobsters, among other species.

Unlike fishermen who catch fish with a swimming bladder that can be seen on an echo sounder, the Norway lobster fisherman has no way of knowing whether the Norway lobsters are there to be caught. Norway lobsters live in small caves under the seabed and can only be caught while they are up on the seabed.

Fishermen have many years of experience which can give an indication of where and when it is best to fish for Norway lobsters. But it is only an indication. And it is based on what is in the trawl when it is pulled up after five to six hours in which it has dragged along the seabed for approximately 25 km.

“The fishermen lift it up and assess: Is this good?  If it is, they continue doing more or less the same thing. Perhaps the whole catch was landed within 45 minutes, but they don't know where,” says Ludvig Ahm Krag.

The new camera system will enable the fisherman to react when the live video shows whether the trawl is catching the right species and sizes.

“This technology has the potential to transform fisheries into being more about precision fishing. This will reduce the impact on the seabed as well as CO2 emissions—because they will interrupt the trawling when there are no Norway lobsters—and the fisherman will get more value out of the quota he has,” says Ludvig Ahm Krag.

Competences and collaboration

The system is a result of several years of development. It consists first and foremost of a camera, which is positioned where the fish are entering the catch bag, with a cable that sends the signal directly to the fisherman’s wheel house.

Researchers and technicians from DTU Aqua have developed the first system in which the camera angle and light—in interaction with the right background—provide extremely clear images of the catch. Among other features, a tarpaulin design has been added to the trawl (see visualization) which prevents seabed sediment from muddying the water and making the images unclear. This has previously made use of this type of camera footage impossible.

“The technicians from DTU Aqua’s monitoring office have developed and built components so that we could examine whether it was at all possible to send a clear and stable video signal from a trawl. We have been able to test the equipment and gradually adjust it on our research ship Havfisken, which is rigged as a cutter and manned with people with experience in commercial fisheries. But we were also aware that we had to involve a company if it was to become a finished product,” says Ludvig Ahm Krag.

Pooling know-how and ideas

The company Atlas Maridan came onboard as a project partner. Atlas Maridan is a specialist in system design of autonomous underwater vehicles and is part of the German company Atlas Elektronik, which works with underwater acoustics. Atlas Maridan had already had its eye on the fishing industry for some time, says Senior Engineer Max Abildgaard (dr.ing.):

“We originally worked with a small compact underwater drone with a camera and artificial intelligence (AI) for species recognition of fish. But in our consultations with the fishing industry, it became clear that they would prefer having the camera and AI solution, but without the drone. So the focus was changed and aimed at designing an underwater camera.”

One of the company’s industrial PhDs got in touch with Ludvig Ahm Krag.

“It was fairly quickly clear to us that DTU had technology, knowledge, know-how, and, not least, contacts in the environment. They also have a good understanding of where to position light and camera in trawl. So it was a quite obvious choice for us to pool our know-how and ideas. I’m really looking forward to continuing our collaboration and to marketing the solution,” says Max Abildgaard.

From hardware to software that can keep an eye on the catch

Atlas Maridan has now produced a solution that has been demonstrated to fishermen and is ready for sale and installation on the cutters. Ludvig Ahm Krag expects that the system has the potential to be used in commercial fisheries worldwide.

The camera hardware contains the sum of the partners’ ideas and competences. The next step in the collaboration is to complete the software development, so that a computer can process data using artificial intelligence.

A PhD student in DTU Aqua’s fisheries technology group has developed an algorithm in collaboration with researchers from DTU Aqua’s observation technology group, so that a computer can recognize and count Norway lobsters. This feature is currently being further developed in collaboration with DTU Electrical Engineering, so that the computer can also recognize fish species.

The goal of this work is that the fisherman does not have to keep a constant eye on the video signal because the computer will be able to process data and, for example, sound the alarm. This could be the case if there are too many small cod in the trawl, or if there are no Norway lobsters at all.

Is being tested on two commercial vessels

Two cutters are currently having the equipment installed in an additional project to look at the effect in connection with commercial fisheries of deep-sea shrimp and Norway lobsters.

One of the cutters belongs to Peter Husth, who has 20 years of experience in Norway lobster fishing:

“I believe it could make a difference. There will be less waste of time, and it should also save fuel.

We do many trawls for five to six hours, where we try to see whether there are Norway lobsters around on the day in question. If we can see right away that there are none, then there is no reason to continue fishing there.

I don’t think anyone has completely figured out the patterns of Norway lobsters. In one trawl you can get 500 kilos, and then when you lower the trawl again—you get none at all. It will be fun to see what it looks like on the seabed when there are large shoals. They must be virtually crawling on top of each other.”