【Algomatic】In a joint application with Mitsubishi Heavy Industries for the GENIAC-PRIZE hosted by METI and NEDO, won 2nd place in the theme of 'Formalizing Tacit Knowledge in Manufacturing'.
Algomatic Inc., in collaboration with Mitsubishi Heavy Industries, Ltd., has won second place in the GENIAC-PRIZE, a program sponsored by the Ministry of Economy, Trade and Industry (METI) and NEDO to promote R&D and social implementation in the generative AI field. Their winning proposal, under the theme 'Formalizing Tacit Knowledge in Manufacturing,' was titled 'Formalizing Tacit Knowledge through a 'Comparative' Approach of Work Videos of Experts and Non-experts, using TIG Welding as an example.' The project proposes a system to structurally extract and formalize difficult-to-verbalize physical knowledge from manufacturing sites for accumulation and utilization.
📋 Article Processing Timeline
- 📰 Published: March 26, 2026 at 22:56
- 🔍 Collected: March 28, 2026 at 21:59 (47h 3m after Published)
- 🤖 AI Analyzed: April 14, 2026 at 22:48 (408h 49m after Collected)
Algomatic Inc. (Headquarters: Minato-ku, Tokyo; CEO: Shunsuke Ohno) is pleased to announce that it has won second place in the 'Formalizing Tacit Knowledge in Manufacturing' theme of the 'GENIAC-PRIZE,' a program by the New Energy and Industrial Technology Development Organization (NEDO) aimed at promoting research, development, and social implementation in the field of generative AI. The entry was a joint submission with Mitsubishi Heavy Industries, Ltd. (Headquarters: Chiyoda-ku, Tokyo; President & CEO: Eisaku Ito). The proposal, titled 'Formalizing Tacit Knowledge through a 'Comparative' Approach of Work Videos of Experts and Non-experts, using TIG Welding as an example,' presented an initiative to structurally extract, accumulate, and utilize physical knowledge that is difficult to verbalize in a manufacturing environment. (Left: Mr. Kazuya Tsutsumi, Chief Research Officer, Mitsubishi Heavy Industries, Ltd. Research & Innovation Center; Right: Yuki Nanri, CTO, Algomatic Inc.)
◾️About GENIAC-PRIZE
GENIAC-PRIZE is a NEDO Challenge program with a total prize pool of approximately 800 million yen, targeting four themes across three domains where solutions using generative AI services are desired: social issues, government administration, and safety. Led by METI and NEDO, it was held to encourage development and demonstration by diverse entities from various regions, aiming to accelerate the utilization and social implementation of generative AI applications. At the final screening and awards ceremony on March 24, 2026, a total of 42 prize winners were selected from over 200 applications.
https://geniac-prize.nedo.go.jp/
【Domain Themes】
Social Issues: Development of AI agents to solve social issues using domestic foundational models.
Theme I: Formalizing tacit knowledge in the manufacturing industry.
Theme II: Improving productivity in customer support.
Government Administration: Development of generative AI to contribute to the efficiency of examination work in government offices (Theme III).
Safety: Development of risk exploration and risk reduction technologies to ensure the safety of generative AI (Theme IV).
【YouTube】GENIAC-PRIZE Final Screening and Awards Ceremony ※The presentation by Mitsubishi Heavy Industries and Algomatic begins around 3:10:56.
◾️Proposal by Mitsubishi Heavy Industries and Algomatic
Proposal Title: Formalizing Tacit Knowledge through a 'Comparative' Approach of Work Videos of Experts and Non-experts, using TIG Welding as an example.
Proposal Overview: In this proposal, by simply recording and uploading work videos of an expert and a non-expert, an AI automatically analyzes the differences in their work and visualizes the tacit knowledge. Depending on the target task (in this case, TIG welding, where quality standardization is difficult), the AI selects the optimal method from multiple analysis modules to extract the 'differences' between the two videos from various perspectives. This makes it possible to capture skills, including 'physical knowledge' that is difficult for the workers themselves to verbalize, as formal knowledge through comparison. Furthermore, by organizing the extracted differences into a report, it enables technical evaluation and feedback for the non-expert. This is expected to significantly reduce the time and cost of personnel training and contribute to solving challenges in skill succession.
◾️Excerpts from the Presentation Materials
◾️Representative Comments
・Mr. Kazuya Tsutsumi, Chief Research Officer, Mitsubishi Heavy Industries, Ltd. Research & Innovation Center / Ph.D. in Engineering / Professional Engineer (Mechanical Engineering)
'I am greatly honored that our initiative has been recognized and received this award. First, I would like to express my sincere gratitude to Algomatic for achieving this groundbreaking technological development. In the manufacturing industry, there are many cases where the skills and judgments of experts are accumulated without being verbalized, making skill succession and standardization a long-standing challenge. In this proposal, we took on a new approach of extracting and structuring tacit knowledge from the differences by 'comparing' the work of experts and non-experts. I believe this is a step towards visualizing the physical knowledge that could not be captured by conventional formalization methods. We will continue to contribute to the advancement of skill succession and productivity improvement in manufacturing sites, with an eye toward the social implementation of such technologies.'
・Yuki Nanri, CTO, Algomatic Inc.
'I am very pleased that our joint proposal with Mitsubishi Heavy Industries has been recognized and resulted in this award. At Algomatic, we emphasize not just using generative AI as a tool for efficiency, but creating new value by fusing it with human knowledge and experience. This proposal embodies that philosophy, attempting to structure and treat difficult-to-verbalize physical knowledge as reusable intelligence through comparison and analysis. In addition, we are also focusing on operational design that will continue to be used in the field. Especially in the manufacturing industry, where new complex systems...'
◾️About GENIAC-PRIZE
GENIAC-PRIZE is a NEDO Challenge program with a total prize pool of approximately 800 million yen, targeting four themes across three domains where solutions using generative AI services are desired: social issues, government administration, and safety. Led by METI and NEDO, it was held to encourage development and demonstration by diverse entities from various regions, aiming to accelerate the utilization and social implementation of generative AI applications. At the final screening and awards ceremony on March 24, 2026, a total of 42 prize winners were selected from over 200 applications.
https://geniac-prize.nedo.go.jp/
【Domain Themes】
Social Issues: Development of AI agents to solve social issues using domestic foundational models.
Theme I: Formalizing tacit knowledge in the manufacturing industry.
Theme II: Improving productivity in customer support.
Government Administration: Development of generative AI to contribute to the efficiency of examination work in government offices (Theme III).
Safety: Development of risk exploration and risk reduction technologies to ensure the safety of generative AI (Theme IV).
【YouTube】GENIAC-PRIZE Final Screening and Awards Ceremony ※The presentation by Mitsubishi Heavy Industries and Algomatic begins around 3:10:56.
◾️Proposal by Mitsubishi Heavy Industries and Algomatic
Proposal Title: Formalizing Tacit Knowledge through a 'Comparative' Approach of Work Videos of Experts and Non-experts, using TIG Welding as an example.
Proposal Overview: In this proposal, by simply recording and uploading work videos of an expert and a non-expert, an AI automatically analyzes the differences in their work and visualizes the tacit knowledge. Depending on the target task (in this case, TIG welding, where quality standardization is difficult), the AI selects the optimal method from multiple analysis modules to extract the 'differences' between the two videos from various perspectives. This makes it possible to capture skills, including 'physical knowledge' that is difficult for the workers themselves to verbalize, as formal knowledge through comparison. Furthermore, by organizing the extracted differences into a report, it enables technical evaluation and feedback for the non-expert. This is expected to significantly reduce the time and cost of personnel training and contribute to solving challenges in skill succession.
◾️Excerpts from the Presentation Materials
◾️Representative Comments
・Mr. Kazuya Tsutsumi, Chief Research Officer, Mitsubishi Heavy Industries, Ltd. Research & Innovation Center / Ph.D. in Engineering / Professional Engineer (Mechanical Engineering)
'I am greatly honored that our initiative has been recognized and received this award. First, I would like to express my sincere gratitude to Algomatic for achieving this groundbreaking technological development. In the manufacturing industry, there are many cases where the skills and judgments of experts are accumulated without being verbalized, making skill succession and standardization a long-standing challenge. In this proposal, we took on a new approach of extracting and structuring tacit knowledge from the differences by 'comparing' the work of experts and non-experts. I believe this is a step towards visualizing the physical knowledge that could not be captured by conventional formalization methods. We will continue to contribute to the advancement of skill succession and productivity improvement in manufacturing sites, with an eye toward the social implementation of such technologies.'
・Yuki Nanri, CTO, Algomatic Inc.
'I am very pleased that our joint proposal with Mitsubishi Heavy Industries has been recognized and resulted in this award. At Algomatic, we emphasize not just using generative AI as a tool for efficiency, but creating new value by fusing it with human knowledge and experience. This proposal embodies that philosophy, attempting to structure and treat difficult-to-verbalize physical knowledge as reusable intelligence through comparison and analysis. In addition, we are also focusing on operational design that will continue to be used in the field. Especially in the manufacturing industry, where new complex systems...'