World Aquaculture Singapore 2026

June 2-5, 2026

Singapore

AI-BASED PREDICTION OF DISEASE OUTBREAKS IN Pangasianodon hypophthalmus PONDS USING A BEHAVIOURAL AND ENVIRONMENTAL MONITORING SYSTEM

Sunil Kadri*, Tran Minh Phu & Stefan Sawynok

Research Institute for Northern Agriculture, Charles Darwin University

Brinkin, NT 0810, Australia

*sunil.kadri@cdu.edu.au



Vietnam’s striped catfish Pangasianodon hypophthalmus industry is suffering more frequent disease outbreaks associated with intensification and climate-driven increases in water temperature. Current monitoring methods rely on visual observation, appetite changes, or mortality events, meaning that interventions are often deployed only after outbreaks have begun.

Following previous presentations at WAS conferences on AI-driven disease prediction in salmon aquaculture systems using hydroacoustics (Kadri & Kvam 2023, 2024); a project consortium consisting of Charles Darwin University, Can Tho University, Infofish Australia, Vinh Hoan & Viet UC will develop and commercialise a novel platform for predicting disease outbreaks in tropical pond aquaculture. The system integrates hydroacoustic monitoring of fish behaviour with environmental sensing and machine learning analysis to provide early warnings of emerging health risks in striped catfish farming; aiming to shift farm management from reactive disease treatment to proactive prevention through continuous monitoring - irrespective of ambient light levels or water clarity - and predictive analytics.

The system will be deployed on commercial catfish farms in the Mekong Delta, where sensor networks will collect 24/7 behavioural and environmental data across multiple production cycles. The resulting dataset will be used to train predictive models and evaluate their capacity to forecast disease events. The platform combines two main data streams collected within ponds: Hydroacoustic monitoring to detect behavioural patterns and biomass distribution & water-quality measurements. Machine learning algorithms will integrate these data to identify combinations of behavioural and environmental markers associated with elevated disease risk. These markers will be calibrated through repeated production cycles incorporating farm metadata, providing for model refinement prior to commercialisation.

As far as we are aware, this work will represent the first integration of hydroacoustic behavioural monitoring and environmental sensing for disease prediction in tropical pond aquaculture. The resultant technology will be designed as a scalable “plug-and-play” system suitable for both large commercial farms and smaller producers, with potential future application across multiple aquaculture species.

This project will also empower remote, marginalized communities by providing accessible, high-tech aquaculture solutions and co-designed training, enabling them to manage farms efficiently, anticipate environmental and health challenges, and actively benefit from sustainable technological advancement. It aims to reduce isolation and promote equitable participation in Vietnam’s aquaculture growth.

Kadri, S & Kvam, J. (2023) Monitoring whole population welfare status in cages using hydroacoustics. World Aquaculture 2023 “Supporting Strength in Aquaculture”, May 29-June 1 Darwin, Australia. p. 146 https://wasblobstorage.blob.core.windows.net/meeting-abstracts/WA2023AbstractBook.pdf

Kadri, S & Kvam, J. (2024) ​​​​Welfare scoring for​ whole fish populations based on behaviour, production and environmental variables using artificial intelligence. Asia Pacific Aquaculture 2024 “Aquaculture – Driving the Blue Economy”, July 2-5 Surabaya, Indonesia. p. 158 https://wasblobstorage.blob.core.windows.net/meeting-abstracts/APA2024AbstractBook.pdf