In salmon farming, the devil is in the details. Data-driven knowledge is starting to help farmers understand these details better – and act on them.
“Digitalization does not aim replace our colleagues at the farms. Their experience will still be the backbone of this industry. Instead, digital tools help them to make better decisions. Valuable experience is coupled with data-driven facts,” says Trond Kathenes.
He is Grieg Seafood’s Chief Digital Officer, and is as passionate about data as he is about fish. For years, he has advocated for how digitalization, or “Precision Farming” as we call it, can help improve the aquaculture industry. Grieg Seafood have invested heavily in digital infrastructure at all our farms. The foundation had to be built, brick by brick.
Analyses support reduced time at sea
Trond and his team of tech-heads have built Grieg Seafood Analyse hub, where they conduct data and regression analyses together with colleagues that are specialists on biology. The hub provides new insight and tools for strategic and tactical decision support.
Since 2019, they have obtained results with high operational value.
“We conducted data analyses on some of our biggest biological challenges: outbreaks of ISA and winter ulcers in Finnmark. The findings have triggered changes to production to reduce risk of ISA and winter ulcers. This is only one example of analyses that can provide strategic data-driven decision support,” says Trond.
The findings from the data analyses have triggered changes to production to reduce risk of ISA and winter ulcers.
Several of the analyses support reducing the time the salmon spend at sea as a strategic way forward for the industry. After a certain number of days at sea, the likelihood of outbreaks increases significantly, according to the data. Grieg Seafood is already focusing on post-smolt transfer and reduced time at sea as a solution for more sustainable growth and production.
“Understanding the best way to produce post-smolt on land, to optimize fish health and welfare, robustness, and growth, is also a focus area for our analyses. There is much to learn about temperature, light conditions, salinity, and other elements in the freshwater phase, and how these parameters must be adjusted throughout production. Possibly, things must be done differently depending on what time of the year the smolt will be transferred to sea,” Trond explains enthusiastically.
New data analyses are in progress, to gain new strategic insight into other production areas. For example, one project will analyze aspects of how on-growth, sea lice exposure and fish health is impacted by different feed types.
Data-driven support to increase survival
In British Columbia, Grieg Seafood has reaped the benefits of data-driven decision support for some time. While low oxygen levels and harmful algae blooms used to have detrimental effects on fish health and welfare, production and costs, the local Grieg BC team is increasingly successful at mitigating their impact.
“On every farm in the region, we have digital sensors monitoring harmful algae blooms and low oxygen levels, as well as an aeration system that upwells clean, oxygenated water to mitigate events,” says Dean Trethewey, Director for Seawater Production, Regulatory and Certifications.
While such conditions used to be discovered only when it was too late and the fish already impacted, Dean and his colleagues have come a long way in early detection. Going forward, they will improve their ability to predict exposure to harmful algae blooms and low oxygen levels three days in advance. However, to have operational value from data-driven decision support, effective mitigating measures must be available. Grieg Seafood BC is working to further fine-tune that aspect, particularly at sites with the highest exposure.
“In 2020, we trialed adding a 15-meter deep tarp around farms at two high risk sites to get improved effect of our upwelling system. We significantly improved survival rates and growth. Mortality in Grieg Seafood BC related to harmful algae blooms was reduced from 3.4% in 2019 to just 0.9% in 2020,” says Dean.
Grieg Seafood British Columbia’s average survival rate increased from 88% in 2019 to 90% in 2020. This year, they are taking the technology to the next level.
“We will test using nano-bubble technology, to better mix the oxygen into the water column and improve the system further. It is very satisfying for our team to see our work over the past year resulting in better fish welfare – that is precisely what we are aiming for,” Dean states.
Mortality in Grieg Seafood BC related to harmful algae blooms was reduced from 3.4% in 2019 to just 0.9% in 2020.
Learning and benchmarking
Going forward, Grieg Seafood will expand data-driven decision support at the strategic, tactical, and operational level.
“Our dashboards are making it easier to get a good overview of important parameters at each farm, and to compare, benchmark, and transfer best-practices continuously. These tools make it easier to get the full overview of the entire production cycle, and implement lessons learned,” says Trond.
In Rogaland, a full-scale integrated operational center was completed in 2020. The center monitors and controls the production on all farms. New features will be continuously added to the unit. For example, the Rogaland team is conducting pilot trials utilizing video analytics with machine learning algorithms to calculate biomass size and weight, as well as automatic sea lice counting. This technology will be expanded to new areas such as behavioral based fish welfare monitoring. Other initiatives are utilizing Augmented Reality technology.
And this is only the beginning, according to Trond.
“We are only at the start of the digital transformation in our industry. It is very exciting to see how our internal culture is already changing. Colleagues are eager to test out new ways to improve and add to their experience. After all, we all share the same goals: increased growth and productivity, reduced environmental impact, improved fish welfare, and lower costs.”