An IT revolution

Google it
Tidal emerged as one of the spin-offs from the “Moonshot Factory” set up by Alphabet, the parent company of Google, to explore ways in which technology and particularly AI could be applied to aquaculture.

Now a tech business in its own right, it has developed a system of cameras, sensors and machine perception tools that can continuously see, sense and survive in the harsh ocean environment.

Kira Smiley, Director of Sales and Marketing with Tidal, says: “Today, we can use video footage captured from our camera systems to detect and provide daily estimates for biomass, various fish welfare indicators and sea lice levels. In addition, using the live video footage collected, this visual data can be analysed in real-time to allow for autonomous feeding control, where ML [machine learning] and AI systems can use visual indicators such as feed pellets and fish behaviour to change or stop feed rate in real time to enable efficient feeding and reduced feed waste.

“Given that there is now so much video footage collected (eg at Tidal we have tens of millions of fish observed and tens of billions of data points captured), we can now start to analyse this for other capabilities such as more nuanced early indicators for welfare challenges, for example, health issues with visual indicators connected to fish behaviour.

“In addition, with so much real-time data available, this unlocks a robust dataset that will allow technology providers to move from real-time insights to prediction, which would help farmers better plan and prepare for operations. In addition, when you can start incorporating insights up and down the value chain, this unlocks huge potential for genetics and breeding, for validating new sustainable feed performance, and generally for a multitude of services ancillary to aquaculture production.”

Developing this kind of technology and taking it to the next level requires a combination of disciplines, Smiley says. AI and machine learning are important, but so is data science, which is needed to help make sense of the enormous amount of information collected. Advances in complementary technologies such as Tidal’s underwater cameras are also a necessary part of the picture.

Sharing is caring
As the biggest producer in the world’s highest value aquaculture market, it’s not surprising that Norway is the location for much of the cutting-edge technology in this field.

Collecting data is one thing, but data shared throughout the industry is exceptionally powerful.
AquaCloud was established in 2017 and is a big data project anchored in the aquaculture industry’s need to solve common challenges in order to create sustainable growth.

The project is part of NCE Seafood Innovation and began together with cluster members Lerøy Seafood Group ASA, Grieg Seafood ASA, Mowi ASA, Bremnes Seashore AS, Lingalaks AS, Eide Fjordbruk, and Bolaks AS. The project has developed substantially since 2017, and today it involves an even broader group of leading aquaculture companies.

The initial scope of the initiative was to establish a secure database for storing data and to use advanced analytics to identify where sea lice outbreaks were probable. Despite some success, data quality and dependability were insufficient to reach its ambitious goals at the time.

At the core of the project is still the data platform which receives continuous updates from participating companies and represents a unique repository of high-resolution data from their farming operations. Within legal and competitive limitations, selected datasets are shared both between participants and even made generally available to third parties.

The project has evolved from being a pure sea lice forecasting asset to become a hub of industry activities.

AquaCloud is financed through partner contributions, Innovation Norway, and since 2019 also by Siva through the financing of Norway’s Ocean Innovation Catapult Centre.

The system uses an open IoT-based standard, allowing equal access to aquaculture sensors and systems. Even so, sharing is always a sensitive issue between businesses which are competing fiercely and are understandably reluctant to give away too much of their inside knowledge.

Speaking at the Seafood Expo Global event last year, AquaCloud General Manager Kristian Blom said: “Our goal is to facilitate the sharing of data… a lot of farmers don’t see ‘what’s in it for me’.

“A lot of people can create reports, but what the industry needs is insights and predictions.”

Blom acknowledged the difficulties, both in terms of standardisation and sensitivity of data, but he said the rewards for the industry are very real. He added: “I’ve seen vets with tears in their eyes, because the data has been so useful for them.”

Challenges for AI
While AI offers many benefits to aquaculture, there are also challenges.

The initial investment in AI technology can be high, making it difficult for smaller farms or those with a product which is less than premium value to adopt. AI models need large amounts of data to be effective, which can be difficult to collect, especially in remote locations. Successful implementation of AI systems also requires specialised expertise.

Despite all this, however, it is undeniable that an AI-driven revolution in aquaculture is well underway, and we can expect to see the pace of change accelerate still further in the near future. 

Webinar: Technology, Remote Monitoring and Artificial Intelligence

19 February 2025, 11:30 - 12:30

In association with our Panel Partner, Krucial, Fish Farmer will be discussing the use of IT to remotely monitor fish behaviour, oxygen, water temperature, feed usage and health and safety issues. How has AI assisted fish farming operations and what technological advances does the future hold?

Register for this FREE event: click here