FUTURE TECHNOLOGY

Digital twins in aquaculture are advancing fast, but remain far from commercial farms

Portugal, 6 April 2026 | The technology promises to optimise production and reduce risks, but the gap between laboratory performance and real-world deployment remains critical

Digital Twin en acuicultura

Aquaculture already has technologies capable of predicting, optimising and automating large parts of production. Yet their real-world application at farm level remains limited. A recent systematic review published in Aquaculture Engineering, analysing 80 studies on decision support systems (DSS) and digital twins, confirms strong growth in research activity since 2023 while highlighting a clear disconnect between technological development and commercial adoption.

The sector is moving towards data-driven production models powered by the Internet of Things (IoT) and artificial intelligence, enabling real-time monitoring, predictive analytics and automated decision-making. Within this shift, digital twins represent the next step: virtual replicas of farms that allow scenario simulation, risk anticipation and production optimisation without directly affecting the physical system.

However, the study identifies a structural bottleneck: fewer than 10% of these solutions have been validated under real farming conditions. While models achieve high accuracy in controlled environments, performance drops significantly in commercial setting due to biological and environmental variability, data quality issues such as sensor drift, biofouling and connectivity failures, and the difficult of generalising models across species, systems and locations.

Despite these limitations, digitalisation is already delivering tangible results in specific areas. Wate quality management accounts for 45% of studies, followed by disease detection and feeding optimisation. In these domains, systems have demonstrated the ability to reduce mortality during critical events by up to 50%, improve feed conversion ratio (FCR) by 10-15%, and significantly reduce labour costs.

By contrast, digital twins remain at early development stages (TRL 3-5), with no large-scale commercial implementations reported. Their deployment requires continuous high-quality data streams and robust validation in highly variable biological systems, which currently limits their use to research and pilot environments.

The study also highlights a key factor shaping adoption: investment decisions are driven primarily by economic returns rather than sustainability benefits. Technologies that reduce operational costs – particularly feed, energy and labour – are gaining traction, while those delivering mainly environmental benefits face greater barriers without policy support.

To bridge the gap between research and practice, the sector must address several structural challenges, including data standardisation, model robustness across farms, long-term real-world validation, and simplification of technological solutions. Without this, digital twins risk remaining a technological promise rather than a practical tool capable of transforming aquaculture.