A Tokyo-based IT venture has developed a system for automatically judging the appetite of cultured fish using artificial intelligence (AI).

Fish farmers in the Shikoku region have introduced the AI system on a test basis, as they expect that reduction of wasteful amounts of feed to their fish can lower farming costs.

The AI system was developed by Umitron K.K., based in Minato Ward, Tokyo.

Under the system, cameras are installed to monitor water surfaces and inside the fish farming preserves to observe the fish’s movements. The AI automatically quantifies the fish’s appetite based on how they gather when fed and the splash of water on the surface.

The company has also developed a device that automatically feeds fish when their appetite increases.

The information can be used along with the system to control feeding times and feed amounts by remote control.

It is necessary to feed juvenile fish in small portions. They may feed as many as 5-10 times a day. Appetite also varies with water temperature and growth stage.

Excessive feeding not only wastes food but also makes the water dirty.

There are businesses that manage more than 100 fish farming preserves, so the burden of feeding is not small.

“Remote management of fish not only reduces costs but also increases work efficiency,” a company spokesperson said.