Tokai Univ., Fujitsu Develop AI to Test Tuna Freshness

The Yomiuri Shimbun

21:00 JST, January 15, 2023

Fujitsu Ltd. and Tokai University have successfully used an artificial intelligence-based ultrasound test to determine the freshness of frozen tuna.

The “non-destructive” testing method is expected to reduce costs for the fishing industry as it allows freshness to be judged without cutting off the tail of the fish for a visual examination of its cross-section.

The two parties aim to put the system into use within two to three years.

Tokai University and Fujitsu found that ultrasound waves with a frequency of about 500 kilohertz are optimal for ultrasonic inspections of frozen tuna.

Based on the waveform reflected by the backbone of tuna specimens, researchers were able to use AI to correctly discern the freshness of frozen tuna 70% to 80% of the time. The two parties are working to further improve accuracy.

Tuna is rapidly frozen on board fishing boats immediately after being caught. Their tails are cut off at port for experts to determine freshness.

With the newly developed testing method, Fujitsu also hopes to be able to detect abnormalities in fish, such as blood clots and tumors.