Graph databases for visualisation and optimization of aquaculture production processes

Gjennom forskningsprosjektet Aquagraph skal Clarify AS, Eide Fjordbruk AS og SINTEF Ocean rydde opp i datakaoset og hjelpe oppdrettere med å fokusere innsatsen der den trengs. Alt for å kunne ta bedre valg til det beste for fisk og forretning. Men hvordan tar du det beste valget for fiskehelse, miljø og bærekraft, samtidig? Oppdrettere står daglig ovenfor vanskelige operative spørsmål, og når hendelser inntreffer, hva er det egentlig som er rotårsaken og hvordan kan vi hindre at det skjer igjen? Dette er sentrale temaer med stor innvirkning på fisk, fortjeneste og bærekraft uansett om det er oppdrett på land, i en fjord eller langt til havs. I Aquagraph handler det om å bruke moderne datateknologi til å strukturere og kunne se sammenhenger mellom alle datakilder som følger fisken. På samme måte som f.eks. Google strukturerer og gjør all informasjon som finnes på internett søkbar, ønsker vi å kunne gjøre det samme med oppdretternes egne data på et mye høyere detaljnivå enn hva som tidligere har vært mulig. Clarify samler gjennom sin dataplattform operasjonelle data fra settefiskanlegg, båter, oppdrettsanlegg og slakteri. Aquagraph bygger videre på dette fundamentet og gjør det mulig å strukturere data på en ny måte, slik at man får en uavbrutt datastrøm med alt fisken har opplevd. På den måten kan man både visualisere data på nye måter og bruke maskinlæring for å finne kilder til problemer og muligheter for forbedring. Eide Fjordbruk stiller med data og bidrar inn som sluttbrukere i utvikling og uttesting av brukergrensesnitt og analyseverktøy. SINTEF Ocean bidrar med forskningskompetanse og lang erfaring innen både teknologi og biologi/oppdrett.

Through the Aquagraph project, Clarify (formerly Searis) have developed and improved their computer science department, gained insight into the odds and ends of the production process of Atlantic salmon, and in the potential value in the data gathered from this process. The introduction and use of graph databases in the Clarify platform is still in it's infancy, but the Aquagraph project have made it's mark and provided Clarify with an excellent starting point for future integration based on the developed aquaculture onthology. SINTEF Ocean has developed new knowledge in the areas of computer science for aquaculture and graph databases. Being able to work closely with fish farmers and gaining access to real data is truly invaluable when conducting research. SINTEF Ocean has presented project results at several conferences both nationally and internationally, and published one research paper in collaboration with Clarify. Eide Fjordbruk has worked closely with Clarify and SINTEF Ocean to provide valuable insight into their aquaculture data, their production process and their challenges. They have also provided inputs on future developments of the Clarify software, thus strengthening the relationship between Eide Fjordbruk and Clarify. Throughout the project period we have seen how the aquaculture industry has become more and more digitalized, how interesting new technologies are introduced and how the industry is uniting to share more data through initiatives like Aquacloud. The potential for innovation and insight offered by combining the Aquagraph solution with the Aquacloud database is huge, and this is something the project consortium wishes to pursue in future collaborations. Digitalization, data science, machine learning and artificial intelligence is gaining more and more traction in the aquaculture domain. The Aquagraph project, and it's results, will likely have a role to play in the future of digital aquaculture.

The aquaculture industry generates enormous amounts of data every day. Data comes from sensors in the fish cages, cameras, boats, feeding centrals, but also from e-mails and notes made by the people working on the production sites as veterinarians, farmers, or managers. Large data sets are challenging to navigate and many interconnections will be difficult to detect using today's solutions. Thus, the potential value of the generated data in terms of improving aquaculture production, is still not fully exploited. The AquaGraph project will create a novel software tool for production optimization and visualisation of complex interconnections in aquaculture data by using graph databases. The project will focus the research activities (RA) towards: RA1: Graph database foundation, RA2: Visualisation of data in graph databases and RA3: Using graph databases to improve aquaculture operations. The developed tool will contribute to digitalization of the aquaculture industry, more optimal production and improved decision support by combining methods from computer science, cybernetics and the precision fish farming concept.