THE application of artificial intelligence (AI) in fish farming is viable based on a pilot deployment of fish feeders in Malaysia.
The pilot testing of the AI-assisted technology was made possible with a grant from the Grants for Research towards Agricultural Innovative Solutions or GRAINS of the Southeast Asian Regional Center for Graduate Study and Agriculture (SEARCA) based in University of the Philippines Los Baños.
Particularly in Negeri Sembilan in Malaysia, the installation of automated feeders in fish tanks has reduced time spent on manual feeding by 70 percent and cut feed waste by 20 percent.
SEARCA said that for fish farmers, this means saving at least six hours of labor each week and earning more from better feeding efficiency.
"This project has effectively showcased the power of integrating cutting-edge technologies to drive positive change in the aquaculture sector, paving the way for a more sustainable, efficient, and prosperous future for small-scale tilapia farmers," said Ahmad Fikri Abdullah from Universiti Putra Malaysia in Selangor.
SEARCA also said that by using AI-based computer vision to estimate fish biomass, the fish feeder automatically dispenses feed in precise amounts.
Also, the feeder is fitted with Internet of Things or IoT sensors to monitor water conditions like pH, ammonia and oxygen levels, enabling smart, real-time decision-making for optimal fish growth.
One of the farmers who tested the AI-assisted fish feeder, Khairi from Kuala Pilah 101 kilometers (km) from Kuala Lumpur, said it makes the work easier, and speeds up fish growth. He grows tilapia fingerlings in round fiberglass tanks measuring seven feet in diameter and sells them across Peninsular Malaysia.
Fikri's team developed the fish feeder and adjusted feed sizes to farmer requirements with bigger pellets for grow-out tilapia.
Another farmer named Fendi from Port Dickson 84 km from Kuala Lumpur, who fulfills advance orders for grow-out tilapia, has the same feeder installed in his rectangular cement tank.
"It helps to improve feeding efficiency, reduce waste, and monitor in real time. Moreover, it enables the farmer to make accurate decisions and reduce operational costs," he said, sharing the benefits of the system.
SEARCA noted that what sets the system apart is its accessibility. The feeder was built using locally sourced materials, including repurposed water bottles for hoppers and 3D-printed parts for the dispensing unit. The device is both cost-effective and reliable.
Fikri's team intends to monitor and quantify the long-term economic returns for farmers. Building on the early success of the smart fish feeder, the team is planning to add features like disease detection and automated water quality control, turning the feeder into a comprehensive farm management platform.
"The future of aquaculture is not just about bigger harvests. The AI-based automated fish feeder is about smarter systems that work for the people who need them most," SEARCA said.