Tokai University and Fujitsu said they have developed new technology to check the freshness of frozen tuna in Japan. The joint research focused on the development of a novel ultrasonic AI technology, which resulted in the world’s first technique capable of measuring the meat quality of frozen tuna without the need to cut or damage the product.
Demand for tuna has increased significantly both in Japan and globally, with 15 countries fishing and producing more than 50,000 tons of tuna in 2020. The recent global Japanese food boom has increased demand for high-quality tuna, which is mainly used for sashimi (raw fish).
Most wild-caught, natural tuna are quick-frozen aboard commercial fishing boats and then shipped to consumers via traders to restaurants and supermarkets. However, the quality of the tuna largely depends on the conditions at the time of catch and the way it is treated throughout the distribution process.
Conventional methods of inspecting the freshness and meat quality of frozen tuna usually require inspectors to cut off the fish’s tail in order to visually examine a cross-section of the tuna’s tail. Tail cutting off the tuna often damages and ultimately reduces the value of the fish, and the process relies heavily on a limited number of experts trained to accurately perform quality checks.
Ultrasonic waves are used as a non-destructive testing method in quality testing in various areas. However, using frozen products such as tuna proved difficult due to the high attenuation of sound waves.
To address these issues, Tokai University, led by Professor Keiichi Goto, Department of Fisheries, School of Marine Science and Technology, and Fujitsu conducted joint research to examine frozen tuna with low-frequency, low-attenuation ultrasonic waves to determine the freshness of the fish to examine fish. By analyzing the waveforms using machine learning, the two parties successfully developed the world’s first method of determining the freshness of frozen tuna without the need to cut the product.
To find the optimal ultrasonic frequency for examining frozen tuna, Tokai University and Fujitsu conducted experiments at multiple wave frequencies. Tests showed that ultrasonic waves with a relatively low frequency of around 500 kHz gave optimal results.
In order to determine possible indicators of insufficient freshness, the two parties compared ultrasonic waveforms of tuna samples of good and insufficient freshness quality to investigate whether the waveforms differed depending on the freshness of the samples. As a result, Tokai University and Fujitsu discovered that the reflection intensity was particularly intense in the central bone portion of tuna samples with insufficient freshness. Based on these findings, the two parties created a machine learning model based on reflected waves from the middle bone of the tuna samples, capable of correctly inspecting the freshness of frozen tuna with an accuracy of 70% to 80%.
In addition to waveforms that can be easily distinguished by the human eye, the newly developed AI technology is also able to detect differences in waveforms that are difficult to visually perceive.
Possible application scenarios for the new technology
© JCN Newswire