The South Korean government is implementing a four-phase strategy to expand the transition to artificial intelligence (AI) in manufacturing from individual factories to supply chains and industrial complexes. The plan aims to create a manufacturing AI ecosystem where existing AI models can be reused across different processes and utilized collaboratively by large corporations, partners, and industrial complexes.
On November 3, the Ministry of Trade, Industry and Energy held an 'AI Factory Strategy Meeting' at the Korea Press Center in Seoul, where they announced the 'AI Factory Expansion Strategy.' The meeting was attended by over 200 business leaders and experts, including Minister of Trade, Industry and Energy Kim Jeong-kwan, Hyundai Mobis CEO Lee Gyu-seok, HD Korea Shipbuilding & Offshore Engineering President Kim Hyung-kwan, LS Mtron CEO Shin Jae-ho, and Seoul National University College of Engineering Dean Kim Young-o.
According to the ministry, the government has supported the transition of 245 companies to AI factories so far. An analysis of 38 exemplary cases revealed an average productivity increase of 30.1% and a 15.5% reduction in defect rates.
However, the government identified limitations in the duplication of investments and the need to redevelop AI technologies created in individual factories for application in others. To address this, the ministry has developed a four-phase strategy to expand manufacturing AI innovation from individual factories ('points') to similar processes ('lines'), then to supply chains and industrial complexes ('planes'), and finally to a fully integrated AI-operated factory ('three-dimensional').
Initially, the ministry will conduct a detailed analysis of 541 existing AI models to select tasks with high potential for expansion. They will create a visual 'M.AX Core Task Heatmap' and introduce customized prescriptions based on the level of AI transition for each company, along with a return on investment (ROI) evaluation system.
Subsequently, verified AI models will be applied to similar processes. For example, a temperature prediction model for steel heating furnaces could be utilized in another company's heat treatment process. Existing models will be further trained to fit on-site conditions, reducing development costs and time while fostering specialized manufacturing AI companies.
The ministry will also promote expansion at the supply chain and industrial complex levels. Large corporations will develop AI models to solve common manufacturing issues in collaboration with partners, identifying shared needs in regionally specialized industrial complexes such as automotive and machinery in Gwangju and Changwon, semiconductors and displays in Cheonan, Cheongju, and Gumi, and petrochemicals and steel in Ulsan and Yeosu.
Ultimately, the goal is to implement a 'full-stack AI factory' that integrates AI control from production planning to manufacturing, inspection, and logistics. The ministry plans to conduct on-site demonstrations and technology development while modularizing and standardizing verified technologies for broader application in other factories.
Earlier, in August, the ministry reported on the achievements of 170 AI factory support sites and set a goal to establish 500 AI factories by 2030. In April, they announced 32 new leading AI factory projects worth 52.75 billion won to expand support.
The ministry will also encourage private investment in the expansion of manufacturing AI. They plan to support outstanding AI companies and large-scale manufacturing AX projects using the Industrial Growth Fund and the National Growth Fund, as well as promote training for current workers and young professionals through the 'M.AX Academy.'
Minister Kim Jeong-kwan emphasized, "We cannot afford to experience trial and error by redeveloping technologies that have already succeeded in other factories from scratch. We must focus on how quickly and broadly we can disseminate verified technologies to translate into tangible sales and profits for companies." He added, "We will not shift the initial risks of introducing and spreading new technologies solely onto businesses and the workforce."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.