FishMet: A Digital Twin Framework for Appetite, Feeding Decisions and Growth in Salmonid Fish

The FishMet framework introduces a digital twin model designed to simulate and predict appetite, feeding behaviour, and growth in salmonid fish, such as Atlantic salmon and rainbow trout. Unlike traditional black-box artificial intelligence models, FishMet integrates neurophysiological feedback mechanisms to realistically model feeding decisions and energy budgets. The system supports real-time data assimilation and scenario testing through a cloud-based application programming interface (API), enabling precision feeding strategies in aquaculture. Validation experiments demonstrated accurate predictions of feed intake and growth, highlighting FishMet’s potential as a decision-support tool for sustainable and efficient fish farming.