Developers: | Dentsu |
Date of the premiere of the system: | July, 2020 |
Branches: | Food industry |
2020: Restaurants use artificial intelligence for determination of quality of a tuna
In the middle of July, 2020 the Japanese network of sushi restaurants began to use the application on the basis of AI for determination of quality of a tuna. It is the key stage in making sushi which is traditionally requiring many years of training.
The application under the name Tuna Scope was developed by the Japanese advertizing company Dentsu. The service uses the algorithms trained at thousands of images with cross cuts of a tail of a tuna. According to the photo the application estimates a tuna on a five-point scale on the basis of visual characteristics, such as gloss of flesh and stratification of fat. For the experienced cook these attributes tell a lot of things about what life conducted fish, than ate and was how active, so, and about expected taste.
Dentsu claims that during tests where the application was compared to experienced people buyers, more than four times from five put down to Tuna Scope the same mark. However, experts in sushi and purchasers of fish estimate capability of Tuna Scope more carefully to replace experienced professionals, especially those which buy meat for high-quality sushi and sashimi.
Nevertheless, developers consider that the application will be very useful to basic quality evaluation of the purchased fish. Besides, the AI-application can be demanded by young people who not always adopt traditional skills at the senior generation.
The Tuna Scope application is already used for assessment of fish in network of Kura Sushi restaurants who offers cheap sushi. This network also uses other devices for economy of costs, for example, robotic dishwashers. Kura Sushi buys 70% of fish for sushi abroad and is afraid of problems with quality of deliveries to pandemic time. The application gives the chance to purchasers of Kura Sushi to estimate fish on places, without waiting for transportation to Japan.[1]