AIFISH in CREAfuturo: A Clearer View of Aquaculture

Underwater imaging, artificial intelligence and marine engineering are bringing better information to fish farming.

Every feeding decision in aquaculture depends on what is happening below the surface. More frequent estimates of fish size and biomass (the total weight of the fish stock) can give farmers a stronger basis for deciding how much feed to provide.

CREAfuturo describes this opportunity in Occhi digitali per l’acquacoltura, with a focus on AIFISH, the project led by Elements Works in collaboration with CREA’s aquaculture researchers. The system combines underwater stereo cameras engineered by Elements Works with artificial intelligence and computer vision developed through this collaboration, to estimate fish size and weight without capture or handling.

For us, this is where the practical value of the project becomes clear. A measurement should help an operator make a decision. In fish farming, that means connecting observations of growth with feeding plans, and making it easier to follow changes over time.

CREA reports weight estimation accuracy reaching 97% and life cycle assessment results indicating a potential reduction of up to 17% in the environmental footprint of farming. These are research results reported by CREA; their application across different farms requires attention to the conditions in which the system operates.

There is an important engineering challenge behind those numbers. An underwater camera must provide useful images in the environment where it will operate, while the software must turn those images into information that farmers can interpret and use. The value of the complete system depends on both parts working together.

This is also why the collaboration matters. Research expertise and marine instrument engineering address different parts of the same problem. Connecting them creates a route from scientific development to tools that can support farm management.

We see a further opportunity in connecting observations of fish growth with measurements of the surrounding water. Temperature, dissolved oxygen and salinity provide context for understanding what is happening in a cage or tank. Bringing these observations together can support a more complete understanding of the farming environment.

Our aim is to make this approach useful and accessible to aquaculture operators, including smaller farms. Better information should support better decisions about feed, fish welfare and the resources needed for production.

We thank the CREA team for the collaboration and for sharing this work through CREAfuturo. AIFISH represents a shared direction: using research and engineering to help aquaculture become more efficient, more informed and more sustainable.