Inheriting the craftsmanship of cultivation technology from data by AI×Fermentation

Fermentative production, Deep Learning, AI

“We want to contribute to raising the level of all industries related to bio-production by advancing technological innovation utilizing biotechnology and digital technology.”

“We want to contribute to raising the level of all industries related to bio-production by advancing technological innovation utilizing biotechnology and digital technology.”
A bio-production management system is being developed with AI that exceeds human abilities in understanding complex cause-and-effect relationships that , based on a totally new productivity indicator using data resulting from various sensors and mutual relations with well-skilled cultivation technology. Leading the construction of a global information base is essential for developing the system to be the backbone of the bioeconomy market.

Background issues

Development of production technology using biotechnology such as fermentation production are recently stagnated, and production bases are gathering at areas where the production is cheaper.
The expectation of Bio-AI and Bio-digital combinations are growing, however, data suitable for the machine learning system have not yet acquired.
In order to develop new production technology, acquiring data that can be utilized for methods such as deep learning is required.

Our technologies

Sensing technologies that replace human senses and capture cultivation conditions undetectable by humans
IoT technologies connecting multidimensional sensing devices to the cloud
Data preprocessing techniques to prepare acquired data for machine learning
Machine learning model development specialized for bio-manufacturing
AI-based automatic control technologies to solve a wide variety of cultivation challenges

Current status

Since June 2020, the project “Development of Production Technology for Bio-based Products to Accelerate the Realization of Carbon Recycling” has been selected as a commissioned project by New Energy and Industrial Technology Development Organization (NEDO).
Since April 2023, operations have begun at a Kyoto University site, running a 30-liter scale biofoundry, establishing it as one of the leading biofoundry sites in the Kansai region.
In September 2023, the results of joint work with KYOWA HAKKO BIO CO.,LTD., Ltd. (improving the production of target substances through AI-based automatic control) were published in the Nikkei (Japan Economics Newspaper).
In December 2024, CHITOSE’s AI technology was set to be applied to Ajinomoto Co., Inc.’s theme, “Development of Precision Fermentation Technology for Environmental Protection and Stable Food Supply,” under the NEDO public project “Research and Development of Technologies to Promote Biomanufacturing.”
For reference, see the press release issued by Ajinomoto Co.: “Ajinomoto Co., Inc. Selected for NEDO Project, Begins Development of New Protein Production Technology”, December 26, 2024
https://news.ajinomoto.co.jp/2024/12/20241226.html (JPN)
In June 2025, the results of joint work with Ajinomoto Co., Inc. (doubling protein production through AI-based automatic control) were published in the Nikkei and The Nikkan Kogyo Shimbun (The Daily Industrial News).
CHITOSE continues research and technology development in collaboration not only with the fermentation industry but also with companies and academic institutions possessing diverse technologies in the broader bio-industry.

Locations

Japan

Partner

Kirin Holdings Company, Limited, Ajinomoto Co., Inc., Amano Enzyme Inc., Mitsubishi Corporation Life Sciences Limited, KNC Laboratories Co., Ltd., and others.

MessagePICK UP PERSON

Tetsushi Kawai
General Manager, Tech & Biz Development Div., Chitose Laboratory Corp.

Fermentation production is an industry built on accumulated know-how. Companies with decades of experience hold deep expertise in certain fields, yet when they take on new projects, they often have to rebuild that knowledge from the ground up. Adding to the challenge, much of this tacit knowledge resides with individuals, making it hard to know whether current approaches are truly optimal. In practice, this means that instead of systematically managing microorganisms, teams often find themselves caught up in day-to-day troubleshooting. To change this, we have been developing a data-driven bio-production management system powered by AI. In our trials, we have shown that AI control can achieve levels of process stability that humans cannot, and even double productivity. By broadly applying AI technology to maximize the performance of industrial strains, and combining it with both breeding technology (CHITOSE’s proprietary disparity mutagenesis technology) and cultivation techniques, we aim to contribute to the expansion of the bioeconomy.