Number of AI-adopted firms to surge from 600 to 6,000 by 2030, alongside deployment of 1,000 manufacturing robots
Government pushes for hazardous process automation and in-kind recognition of manufacturing data to ease initial investment burdens for SMEs
Artificial intelligence (AI) adoption among South Korea's small and medium-sized manufacturing enterprises stands at a mere 1%. While major conglomerates accelerate the development of autonomous manufacturing technologies combining AI and robotics, many small-scale production sites lag behind, failing to even achieve basic digital transformation of their production facilities. Hampered by a shortage of capital and specialized talent—coupled with varying equipment and production methods across factories—applying standardized automation systems remains a daunting challenge.
To narrow this widening gap, the South Korean government plans to channel more than 5 trillion won ($3.6 billion) in national resources into overall smart manufacturing innovation by 2030. The initiative centers on scaling up "physical AI"—where AI assesses production conditions and robots execute physical tasks—across small and medium-sized manufacturers.
The Ministry of SMEs and Startups (MSS) unveiled the "Strategy for Expanding Physical AI in Small and Medium Manufacturing" during the 16th National Policy Coordination Meeting held on the 29th at Chong Kun Dang's Cheonan plant in South Chungcheong Province.
Targeting Hazardous Processes First; Tenfold Increase in AI-Adopting Firms
The government aims to expand the number of AI-adopting companies from approximately 600 to 6,000 by 2030, while raising the digital transformation rate of small and medium-sized manufacturers from 30% to 50%.
Support mechanisms will be tailored to a company's technological maturity. Factories struggling to properly collect production data will be prioritized for foundational data-collection infrastructure, while those already operating smart factories will receive support to apply AI-driven process optimization and digital twin technologies.
Robotics will be deployed first in repetitive manual tasks and high-safety-risk processes. The government plans to support the deployment of manufacturing robots in approximately 1,000 small and medium-sized factories by 2030.
Robots will be assigned to repetitive tasks such as component handling and product transport, while AI analyzes equipment operating status and production data to dynamically adjust operations. While conventional automation focused on repeating fixed motions, physical AI connects decision-making and execution based on real-time shifts in production environments.
Furthermore, the government will promote integrated AI and robotics demonstrations targeting not only individual equipment but entire processes and plants. After validating technology effectiveness in actual production environments, authorities plan to establish scalable models applicable to similar manufacturing sites.
Because small and medium-sized manufacturers feature diverse products and equipment configurations, directly adopting automation systems utilized by conglomerates is difficult. The government intends to establish standardized models so that field-tested technologies can be readily transferred to other enterprises.
Recognizing Factory Data as In-Kind Capital to Lower AI Adoption Costs
Building an AI system requires hardware to collect and analyze data from production equipment. Enterprises must also secure personnel to manage and improve the system post-implementation. For SMEs, both initial capital expenditure and ongoing operational costs present heavy burdens.
The government is pursuing initiatives to evaluate the economic value of manufacturing data accumulated by SMEs and recognize a portion of corporate matching funds in-kind under government support programs.
This framework is designed to allow companies to leverage their existing production data to reduce out-of-pocket cash requirements for participating in support programs. Implementing this scheme will require detailed regulatory frameworks, including data valuation criteria and recognition scopes.
Data standardization will be pursued concurrently to ensure that information collected from disparate factory equipment and systems can be effectively utilized for AI training and applied across various production floors.
The technology supply ecosystem will also be broadened. The government plans to foster 500 specialized smart manufacturing enterprises in robotics, sensors, AI, and digital twins by 2030, providing high-performing firms with comprehensive support spanning R&D, commercial demonstration, and overseas expansion.
The existing Korea AI Manufacturing Platform (KAMP) will be upgraded to "Manufacturing AI 24." This system will establish an end-to-end cycle of collecting manufacturing data, developing AI models, deploying them to actual plants, and continuously accumulating feedback data.
The 5 trillion-plus won announced by the government represents the medium-to-long-term national budget earmarked for overall smart manufacturing innovation through 2030. Specific thresholds for in-kind data recognition and corporate support conditions will be finalized in subsequent follow-up projects.
