Verifying the Scalability of AI Agents Over a Six-Month Period

The Ministry of SMEs and Startups announced on June 29 that it has launched a demonstration project for a ‘Multi AI Agent’ designed to optimize processes and quality in small and medium-sized manufacturing enterprises.


SMEs Ministry Launches Demonstration Project for 'Multi AI Agent' Specialized in SME Manufacturing View original image

The ministry held a kick-off meeting for the “Multi AI Agent Specialized for SMEs Demonstration Project” at a food manufacturing site in North Chungcheong Province, together with 12 selected project implementation organizations.


This meeting marked the first on-site launch for the 12 projects that were finally selected for the Multi AI Agent Development (R&D) Program specialized for SME manufacturing. Around 30 participants attended the event, including representatives from the ministry, the Korea Technology and Information Promotion Agency for SMEs (TIPA), the 12 project implementation organizations, and demand companies.


At the food manufacturing site where the meeting was held, participants shared the difficulties faced by SMEs in the production and quality control process, such as quality variation, adjustment of process conditions, and dependence on skilled workers’ judgment. They discussed how AI agents could support on-site decision-making based on actual manufacturing data.


In particular, attendees agreed that it is necessary for SME sites to adopt AI agents that go beyond simple data analysis to support the judgment of workers and managers, and that are connected to actual work processes.


Additionally, the on-site application scenario at Ijella, a food manufacturing company, was shared. In manufacturing processes where quality variation may occur depending on the condition of raw materials or process variables, AI agents will analyze process and product status, predict quality risks, and verify a structure that supports decision making for both workers and managers.


In the case of Kairos Lab, a beauty manufacturing company, the scenario involves using AI agents throughout the entire cycle of multi-product, small-lot production, from prescription design and quality control to process, regulatory review, and clinical efficacy prediction. Through this, the company aims to validate a structure that improves new product development and production efficiency.


The 12 selected projects will use manufacturing site data over the next six months to validate the operation of AI agents, their field applicability, and the scalability for further R&D. The proof-of-concept (PoC) phase will specifically assess the feasibility of AI agents in key manufacturing tasks and their ability to support decision making for on-site workers or managers.


The ministry, together with TIPA, will support data quality verification, connect with the national GPU (graphics processing unit) infrastructure, and provide dedicated committee members, among other supports. Once the pilot research is completed, the practical applicability in real manufacturing sites and research outcomes will be systematically reviewed through public-oriented evaluations.



Kwon Soonjae, Director of Regional Enterprise Policy at the Ministry of SMEs and Startups, stated, “AI technology is expected to play an important role not only in improving productivity and quality at manufacturing sites, but also in enhancing worker safety and environmental conditions. Through this project, we will validate AI models applicable to SME manufacturing sites and support the expansion of outcomes to major manufacturing sectors such as food, beauty, automotive parts, and metal processing.”


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing