Page 21 - Norway export Seafood, Fishing&Aquaculture 2018-19

would indicate trouble. Fish farmers have Watson using artifi cial intelligence, which “As a result, we have to be more
also used under water TV monitors to see was developed to make medical diagnosis,” technological to be sustainable to control
how fi sh react after feeding to see if they’re says Hukkelås. “Watson has also been used both fi sh welfare and the environment,”
full. Never before, however, have so many with sea lice observations.” says Hukkelås. “The only way to go is the
sensors been combined with complex cybernetic approach.”
mathematical modelling tools to simulate Automated operations will be an obvious
the welfare of one million fi sh. benefi t for offshore farms. Currently there is a Development Licenses
crew of about 3-4 people working on Ocean The Directorate of Fisheries also recently
“What we’re trying to experience is to lift the Farm 1, but it could also be remotely operated. granted conditional approval to another
whole industry to more high tech form of This type of technology would be an attractive digitalized aquaculture concept, Cermaq’s
fi sh farming,” says Hukkelås. export to countries, such as China, which are iFarm, for a development license. The new
increasingly turning to ocean farming to fi ll the technology is based on photo recognition
Elementary, Watson growing need for seafood. of the fi sh using sensors with computer
The end goal of this project is to “close the vision that can recognize salmon based on
loop” from measuring behaviour and then The global population is expected to grow their dot patterns. The technology monitors
letting the machine decide what stimuli to to 8.6 billion people by 2030. At that point, growth, sea lice, disease, lesions and other
give the fi sh. The data could indicate that the world will need 70% more protein than health aspects. It then separates the fi sh that
the fi sh need more food or more oxygen is currently available. Oceans cover more need treatment and those that are ready for
and tell the machine to spread more pellets. than two-thirds of the world’s surface, but harvest based on weight.
It could also predict that a sea lice attack is supply only 2% of food protein for human
imminent in the next week and implement consumption. A key way of increasing the “Digitization and machine learning make it
preventive measures. amount of protein to the growing population possible to monitor each fi sh individually
could be to dramatically increase the without stressing the fi sh,” says Harald
“There have been tests using the IBM production of seafood. Takle, iFarm technology research director
and project manager. “This makes it an
important tool for safeguarding our fi sh and
is a giant leap for fi sh farming.”

Norway has so far approved four other
concepts for development licenses, in
addition to SalMar’s Ocean Farm 1. The
directorate has given a green light to
Nordlaks Oppdrett for Havfarm, MNH
Produksjon’s semi-closed cage Aquatraz,
AkvaDesign’s closed cage concept, and
Marine Harvest’s closed farming oval-
shaped offshore concept the Egg.

SalMar’s farmed salmon.
© SalMar

21 21
seafood, fi shing & aquaculture