The AquaCloud system for combining data like environmental factors, sea lice prevalence and fish health was launched by NCE Seafood in 2017. The first goal of the project was to make a system for sea lice warning, by combining existing data from different salmon producers.
The operation center which could predict sea lice
Trond Kathenes, Chief Digital Officer in Grieg Seafood, is responsible for all digitalization in the company and also a member of the steering group of AquaCloud. AquaCloud works very well. We send data to the database, and we have an API (application programming interface) pulling out datasets every day. If we want, we can display sea lice prediction data for one week ahead on our dashboard, he says. But the sea lice prediction is not yet being fed into the dashboard.
First goal: Sea lice warning
Sea lice warning has been achieved, but only with a 70 percent certainty. For making costly decisions about lice treatment, this might not be good enough for the industry. But to get more accurate predictions, we must improve data quality and standardize methods. Some data sources go automatically into the database, but sea lice counts are done manually: 20 fish from each cage are taken up, lice are counted and fed into the database.
– But 20 fish out of maybe 150 000 in a cage is not enough, even though we are working on standardizing the methods of counting, says Einar Wathne.
Data is being gathered from a wide range of cages every day by staff from Mowi, Lerøy and Grieg Seafood, who are running the project, but also Eide Fjordbruk, Lingalaks, Bolaks and Bremnes Seashore.
– I think an accurate lice prediction will depend on the development of a form of automatic sea lice count. We need to get a system that scans the fish, says Wathne. Several tech-companies are working on ways of resolving this issue.
You are welcome to continue reading this interesting article visiting our Annual Report 2019, clicking here AquaCloud_ppt_16-9