UMITRON develops the world’s first autonomous autonomous fish school appetite analysis system “UMITRON FAI (Fish Appetite Index)”

Umitron Corporation UMITRON develops the world’s first autonomous autonomous fish school appetite analysis system “UMITRON FAI (Fish Appetite Index)” Using machine learning, the appetite of fish is determined in real time at sea. Fully automated feeding. …………………………………………………………………………………………… [Image 1

Key Point Developed “UMITRON FAI (Fish Appetite Index)”, the world’s first real-time fish feeding evaluation system using machine learning. Started implementation for existing customers. Recruiting system cooperation partners for cooperation with large systems such as feeding ships. Participated in AquaNor (no booth exhibition) in Norway from August 20-22 and Japan International Seafood Show in Japan from August 21-23. Create opportunities for business talks with interested parties. UMITRON PTE. LTD. (Location: Singapore, co-founder / managing director Masahiko Yamada, hereafter Umitron) has developed a fish appetite analysis system “UMITRON FAI (Fish Appetite Index)”. “UMITRON FAI” is the world’s first algorithm that automatically evaluates the prey of fish in real time by machine learning. At sea -Automatic accumulation of fish swimming data -Real-time feeding status scoring -Feeding control according to score Is an offshore autonomous service that enables Umitron has begun offering “UMITRON FAI” to existing customers who are already using UMITRON CELL. Umitron will continue to develop these value-added services and provide them to existing customers. Feeding costs in aquaculture Aquaculture is a global growth industry. On the other hand, the optimization of feeding costs, which account for the majority of production costs, is the most important issue in aquaculture management. However, the special environment of aquaculture has had many restrictions on feed management. Difficulties in feeding management at sea Producers use a large number of farmed ginger floating on the ocean to grow fish. Until now, a specific person in charge has gone out to the sea every day to check the state of the fish, and has achieved detailed growth. However, it is difficult to divide the role of each ginger on a ship, and in order to know changes in the state of fish and the sea, a designated person checks the fish continuously on the spot every day, and almost throughout the year. It was necessary to manage the growth without break. Also, long-term offshore work is extremely difficult in dangerous environments such as bad weather, and the working time per ginger must be kept to a minimum. Therefore, until now, it has been necessary for the staff in charge to observe the fish being fed and change the feeding method quickly within a limited time at the offshore site. Determining appetite of fish in real time at sea, fully automatic feeding “UMITRON FAI” is the world’s first algorithm that automatically evaluates the fish prey using machine learning in real time. We are developing with the solution of problems in mind. By automatically scoring the fish feeding status from the feeding video taken with UMITRON CELL, the staff in the field can later review the feeding status based on the score. Establish a feeding strategy from a more objective perspective by observing bird feeding status (time-series data) and feeding status (growth data) among multiple ginger during the growing period. Can do. Feeding improvement plan devised by utilizing “UMITRON FAI” can be immediately reflected in the feeding schedule at UMITRON CELL using a smartphone, so there is no time for offshore work for feeding adjustment and at a short time Trying and changing feed changes is possible. By sharing these feeding information, including employees other than the staff in charge of feeding on site, everyone can share the criteria (score) to stop feeding and improve future feeding plans (amount and timing) Can do. In addition, it is difficult for the producers to keep observing the feeding of the whole ginger all day, but UMITRON FAI will observe and evaluate the prey of the fish instead of the producer, so it is more important such as when the prey falls Growth methods such as direct observation of only special scenes are also possible. Actually, fish appetite changes abruptly depending on environmental conditions such as the growth stage and water temperature, so early detection of wasted food and automatic optimization of feeding in real time by analyzing the state of the fish in real time using UMITRON FAI It also leads to [Image 2

Figure: UMITRON FAI is indexed by combining multiple algorithms, and can be used in consideration of the environment and shooting conditions. Utilizing these systems, “1. Capitalization of in-house knowledge at fisheries sites”, “2. Improvement of working environment by reducing offshore work”, “3. Conservation of sea through early detection of waste food”, and “4. We will support “improving management by optimizing early growth and feeding costs”. Umitron has already implemented a beta version for existing users who use UMITRON CELL, and will continue to update new features as needed. Producers, fishermen’s cooperatives, local governments, etc. who are interested in this initiative should feel free to contact Looking for external system cooperation partners In addition, Umitron has started recruiting system cooperation partners to contribute to more production sites by linking Umitron software such as “UMITRON FAI” to large equipment for aquaculture such as feeding vessels. Umitron will participate in AquaNor held in Norway from August 20-22 and Japan International Seafood Show in Japan from August 21-23, and will have business talks about this initiative. If you are interested, please feel free to contact “” in advance. – About Umitron Umitron is a start-up company that uses technology in growing aquaculture to address the food and environmental problems that humans face in the future. Based in Singapore and Japan, we are developing computer models for sustainable aquaculture using technologies such as IoT, satellite remote sensing, and machine learning. By developing and providing computer models that integrate aquaculture know-how from around the world, we are aiming to implement “sustainable aquaculture on the earth” that is safer and more friendly to people and nature. Contact information Business collaboration: UMITRON service representative Email: Public Relations / Interview: UMITRON Public Relations Office Email:

For more information about this release(Japanese): ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━


You May Like

  • Comment ( 0 )

  • Trackbacks are closed.

  1. No comments yet.