ABB wanted to show how modern, digital technology could simplify and make modern fish farming more efficient. The result was a system that improves accuracy in measurements and provides better animal welfare.

“Last May, ABB participated in a workshop with Norway Royal Salmon (NRS) in Tromsø, together with Microsoft. Over three days, we identified several areas where digital technology can be used to simplify and streamline modern fish farming. The project we chose to start with involves calculating how much salmon is in the pens,” says project manager at ABB, Turid Storaas Nastad.

Fish farmers are required to report weekly to the Norwegian Food Safety Authority on how much fish is in the pens and how much the fish weigh. The usual procedure has been to take 20 fish from each pen, once a week, and measure and weigh them. This has formed a basis for calculating the total. The method is labor-intensive, not particularly accurate, and also not good for the fish.

Automated Measurement

Simply put, the solution that has now been developed consists of a stereo camera that takes images of the salmon, models that can recognize the fish, and models that measure the length between specific points on it. This data is used to calculate the amount of biomass in each pen.

”These measurements are much more accurate than the manual method, and in addition, there is no need to lift and handle the fish outside the pen, so animal welfare is also improved,” continues Storaas Nastad.

ABB has 20 years of experience in delivering electrical and automation solutions to the entire aquaculture value chain and has been responsible for the implementation of the project. Norway Royal Salmon has contributed with knowledge of fish farming, while Microsoft has provided advisory services on the cloud solution and on how images can be interpreted and used as data. In addition, ABB brought Computas on board to fulfill its role as a supplier. ABB is a leading provider of industrial digitalization and has expertise in cloud services, artificial intelligence, machine learning, and image recognition. During periods of high workload, Computas has strengthened ABB’s team in these areas.

”ABB and Computas know each other well after many years of collaboration in various industry segments. It is therefore natural for us to use partners like Computas to complement us in specific niche areas,” says Marius Five Aarset, who is head of digital deliveries at ABB.

Ownership of the Solution

One of the Computas consultants who contributed to the project is senior developer Simen Selseng. He says that Computas was well integrated into the project and that they quickly felt ownership of the solution they were helping to develop. And this despite the fact that a large part of the project was carried out during the period when Norway was in lockdown due to the coronavirus.

”It was a bit special to only work via video without meeting each other in real life until much later in the project. But at the same time, it also shows that we can do almost anything as long as we have good digital tools and a good culture of collaboration,” he says.

Professionally, the project has been a good fit for Computas. The Norwegian IT company has solid experience in system and application development on cloud platforms, and has been working with artificial intelligence since the 1980s.

”We contributed to putting the system and, not least, the machine learning into production. Now it was no longer a prototype that we were developing, but the final system that we worked to make as stable and robust as possible. The system is cloud-based, built on Microsoft Azure. Azure is a mature cloud platform and we have used many of the services offered instead of building them from scratch. This includes brand new technology for machine learning,” he explains.

A Good Collaboration

“The collaboration went very well and we feel that Computas became an integral part of our team. They are flexible to work with and adapt their involvement according to the needs we have and provide the expertise we need” – concludes Five Aarset.