Enhance and GS Caltex Team Up on Manufacturing Tacit Knowledge AI Development, Structuring Skilled Workers' Know-How
TECHWORLD ·
✦ AI Summary
Enhance and GS Caltex announced on the 23rd that they will jointly work on the "Development of an AI Model Based on Manufacturing Tacit Knowledge" project as part of the Ministry of Trade, Industry and Energy's "2026 Manufacturing Tacit Knowledge-Based AI Model Development Project."
Enhance will oversee the overall development as the lead research and development organization, while GS Caltex will participate as the demand company, along with OpenGraph Labs, Chonnam National University, and GIST.
The project will structure equipment, documents, maintenance histories, and the judgment of experienced field engineers into an ontology and a guide chatbot, allowing personnel to review causes, inspection sequences, original maintenance records, and other references before making the final decision.
Enhance and GS Caltex announced on the 23rd that they will jointly work on a task to handle manufacturing tacit knowledge with an AI model. The project, titled "Development of an AI Model Based on Manufacturing Tacit Knowledge," will be pursued as part of the Ministry of Trade, Industry and Energy's "2026 Manufacturing Tacit Knowledge-Based AI Model Development Project."
The task will be carried out with each institution taking on a separate role. Enhance will serve as the lead research and development organization and oversee the overall development, while GS Caltex will participate as the demand company. OpenGraph Labs will handle unstructured document parsing, and Chonnam National University and GIST will participate as joint research institutions.
The core of the project is connecting knowledge that is difficult to document with the decision-making process of skilled workers. Enhance will work to structure the experience and judgment criteria of experienced manufacturing engineers into AI.
Examples of the information to be structured include the first items to check when equipment abnormalities occur and the causes to inspect first depending on equipment condition and operating conditions. The goal is to structure and make usable on-site experience.
GS Caltex aims to organize relationships based on equipment, documents, and maintenance histories. It will compile 832,996 target pieces of equipment and about 2,900 unstructured target documents to organize relationships among equipment, parts, abnormal signs, causes, and measures. In this process, interviews with engineers in the device, machinery, electrical, and instrumentation fields will be used as sources for reflecting knowledge, and the interviews will be used to reflect the sequence for checking equipment condition, criteria for narrowing down causes, and additional considerations for exceptional situations.
The on-site knowledge collected this way will be built in the form of an ontology. Ontology functions to structure relationships among heterogeneous data and knowledge, and it will be used to check related equipment, causes, past maintenance cases, and response methods at the same time when a particular abnormal sign appears.
The knowledge system built will be implemented in the form of a "guide chatbot" for on-site use. The chatbot will be used by personnel through natural-language questions in abnormal equipment or maintenance situations, and the company plans to develop it so that it provides possible causes, inspection sequences, original text from past maintenance records, and original references for judgment such as technical standards.
The AI in this project is not intended to replace on-site personnel and deliver the final conclusion. It is designed so that the guide chatbot presents the necessary information and grounds for judgment, and the person in charge makes the final decision after checking the actual equipment condition and related materials.
According to Enhance, the development environment will be built on GS Caltex's on-premises infrastructure. The scope of information used for the project includes equipment data, maintenance histories, and technical documents, and such information will not be taken outside GS Caltex's internal environment. The system is also designed so that sensitive manufacturing data can be used without being sent outside.
Enhance says the project is focused on converting the experience accumulated by individual skilled workers into organizational knowledge. It plans to prepare for a future generational shift among skilled workers and establish a manufacturing AI foundation that can continue to use existing judgment criteria and problem-solving experience.
Lee Seung-hyun, CEO of Enhance, said the company sees it as important not merely to document skilled workers' experience, but to turn it into knowledge that a company can continue to use. He also said the company will build a system that helps on-site personnel check information and grounds for judgment so they can solve problems efficiently.
Source: TECHWORLD · Kim Seung-gi
Original: https://www.epnc.co.kr/news/articleView.html?idxno=407351
References
This article was produced with the help of an automated content generation algorithm.
Source: TECHWORLD
View originalThis article was summarized and organized by BizCrush based on the original article from TECHWORLD. For exact quotations and full details, please refer to the original article.