The Lab is deploying AI to cut costs and improve production efficiencies in aquaculture.

Fish is a major export for Norway, earning a revenue of $11.3 billion in 2019, as per the Norwegian Seafood Council. Norwegian farmers are increasingly using artificial intelligence (AI) in aquaculture, the farming of fish and other marine organisms. The Norwegian AI Lab has developed AI models that can optimize various operations of this business. (AITrends)

Neural networks for healthy salmon

The Norwegian AI Lab has developed neural networks known as “tiny machine learning” that are a combination of hardware and software running on very low power.

“The pervasiveness of ultra-low-power embedded devices, coupled with the introduction of embedded machine learning frameworks like TensorFlow Lite for Microcontrollers will enable the mass proliferation of AI-powered IoT devices,” says Vijay Janapa Reddi, Associate Professor at Harvard University.

In one application, underwater cameras sense from behavioural changes in the fish whether they are hungry. Optimized feeding using this data can help generate major savings because feed amounts to approximately 40% of fish farming costs.

Telenor Research is developing small computers that use “tiny machine learning” and are smart enough to take operational decisions based on data sensed and transmitted by the cameras underwater and around the farm. This “on-the-spot” management is useful in remote locations which have poor internet connectivity, making it difficult to connect to head office or a cloud network.

The AI system can measure fish sizes, and their count and weight, as well as monitor their health and weight. This is a much safer process compared to manual practices where fish farmers checked the salmon by hand.

Tech companies also active in the field

Technology companies such as Microsoft (NASDAQ: MSFT) and ABB (SWX: ABBN) have partnered to use AI to improve the operations of Norway Royal Salmon (NRS), a leading producer of sustainable salmon, selling about 70,000 tons of salmon every year.

Underwater cameras are deployed to record salmon in submerged fish pens that are floating many kilometers distant from the shore.

AI analyzes the images to calculate salmon counts and measurements. This substantially reduces the workload for NRS.

Overall, the benefits are better fish health, minimizing the environmental impact and a reduction in costs.

In another effort, Norway’s Seafood Innovation Cluster, an industry-funded member organization that aims to foster growth through technology, partnered with IBM (NYSE: IBM) to help the aquaculture industry manage sea lice outbreaks more effectively.

Sea lice are the biggest threat to Norway’s wild salmon population, and to its valuable salmon farming industry.

IBM built AquaCloud—a solution that automatically extracts data from the farmers’ Mercatus and Fishtalk ERP systems on a daily basis and combines it with open data sources provided by the Norwegian government and other sources.

IBM then used an algorithmic model to predict the probability of a sea lice outbreak for each salmon farmer over the next two weeks.

This gives farmers enough time to take defensive action to protect their fish from sea lice.

Related Story:   How AI Watches For Pirates On The High Seas