Integrating artificial intelligence in aquaculture: opportunities, risks, and systemic challenges

Abstract

Aquaculture plays a significant role in the food chain and in rural economies. Continuous water quality monitoring, health management, real-time growth monitoring, and biomass estimation are critical aquaculture activities. Recent advances in computer vision, image processing, and Artificial Intelligence (AI), particularly in Machine Learning (ML) and Deep Learning (DL), enable the control, drive, and solve the problems related to daily real-time aquaculture activities. The performance of various ML and DL models and the quality and availability of public datasets in this domain remain underexplored, and the redefinition of the role of AI in aquaculture is the motivation for this study. This survey aims to analyse and evaluate various methods and datasets for monitoring water quality, estimating fish biomass, disease prediction, and behavioural analysis, to highlight recent developments and to seek the attention of researchers to address the challenges and concerns of aquaculture systems. Currently, there is no single, generalised AI model capable of performing all essential aquaculture activities using continuous time-series data generated from diverse, heterogeneous ponds across a wide geographical area. To address this gap, a 5G-enabled Federated Learning (FL) framework utilising Unmanned Aerial Vehicles (UAV) for cooperative data collection and model training is highly recommended. © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2025.