Aquaculture AI
A KFAS-funded project developing automated machine learning systems for disease prediction and intelligent monitoring in Kuwait's aquaculture sector, in collaboration with the Kuwait Institute of Scientific Research.
Motivation
Aquaculture in Kuwait has become an increasingly important source of fish food production, yet the industry faces persistent challenges from annual disease outbreaks among farmed species. Conventional diagnostic techniques identify disease events too late to mitigate their impact on production yield and profitability, creating economic risk and broader food security concerns. Rotifers — microscopic organisms essential to larval fish feeding — are particularly vulnerable to environmental fluctuations, making their production automation a high-priority target.
This project develops an automated machine learning-based system to anticipate disease onset and dynamically control tank conditions, minimising disease risk while maintaining parameters for optimal production. The work is structured in five stages: (1) automation of rotifer production processes; (2) continuous data collection and environmental monitoring of aquaculture containers; (3) anomaly and early-stage disease detection; (4) enhanced disease characterisation through multi-modal data fusion; and (5) predictive disease prevention with adaptive control.
The system is experimentally validated in aquaculture tanks at the Kuwait Institute of Scientific Research (KISR), leveraging sensors measuring water temperature, salinity, and acidity. The project is funded by the Kuwait Foundation for the Advancement of Sciences (KFAS, grant PN23-14SE-1992) for a three-year term.