"Amid the dual crises of an aging workforce and population decline on the shop floor, AI Transformation is a matter of survival, not choice. Korea's global manufacturing competitiveness can only be safeguarded by advancing beyond simple process automation toward 'autonomous factories' where processes make decisions and prevent defects independently."
Byung-dong Yoon, CEO of ONEPREDICT, emphasized during an interview with E-Daily at the ONEPREDICT office in Banpo-dong, Seoul on the 7th that the future paradigm of manufacturing is shifting from "Automation (Smart Factory)" to "Autonomy (Manufacturing AX)."
As an active professor of Mechanical Engineering at Seoul National University who has long researched the integration of physical systems and AI, Yoon noted, "Past smart factories and today's manufacturing AX differ right from their core goals."
He explained: "While smart factories focused on 'automation' making processes run automatically the goal of manufacturing AX is 'autonomy,' where AI independently evaluates and takes corrective actions across quality inspection, maintenance, and process optimization. Rather than relying on human experience to inspect quality and resolve problems, the structure is evolving toward having AI agents embedded directly within processes to autonomously resolve issues."
<Source : Edaily>
Next After Smart Factory is 'Autonomy'… AI Evaluates Processes and Resolves Problems
Yoon analyzes that manufacturing AX has emerged as a central industry talking point due to the intersection of structural crises on the factory floor and the maturation of AI technology.
"As the working age population shrinks and veteran technicians retire, long accumulated tacit knowledge is vanishing," he said. "Just as urgency was mounting, data accumulated through smart factory initiatives combined with advancing AI technology, joining forces as a solution to this societal issue."
Yoon diagnosed that this situation could actually serve as a golden opportunity for South Korean manufacturing. While the US has already lost a significant portion of its shop floor master technicians, and Germany and Japan face delayed AI transformation due to slower Digital Transformation (DX) despite retaining veteran craftspeople, South Korea sits in a unique sweet spot.
Korea still retains experienced shop floor experts while having already progressed significantly in manufacturing DX.
Yoon evaluated: "Even though Germany and Japan have master technicians, they face a mismatch where digital transformation has lagged, resulting in a lack of data and delayed AI adoption. Korea still has experts with tacit knowledge on the ground, and because DX has progressed to accumulate data, we are in an advantageous position to rapidly embed this knowledge into AI for global competition."
In short, digitizing and embedding the experience and heuristics of skilled workers into AI before they leave the field will determine the competitiveness of Korean manufacturing.
"AI Finds Defect Causes That Humans Miss"… Beyond Predictive Maintenance to Autonomy
A prime domain requiring AI is mysterious quality issues that arise during manufacturing cases where components pass inspection individually, but unexplained defects occur after final assembly. Because multiple processes and countless variables are intricately intertwined, finding root causes through human intuition alone is nearly impossible.
Yoon compared the difference between humans and AI to the Go playing AI 'AlphaGo.'
"No matter how skilled a human expert is, human experience and vision have limits, making it impossible to calculate all complex variables," he stressed. "In contrast, AI can weigh all possibilities probabilistically to pinpoint and solve problems where tens of thousands of complex variables are entangled."
The goal of manufacturing AX does not stop at flagging anomalous signs. True autonomy means discovering anomalies, analyzing root causes, and directly orchestrating optimal responses.
Yoon is implementing the manufacturing AX paradigm he has long advocated on the front lines through ONEPREDICT, the industrial AI company he founded in 2016.
Beyond traditional Predictive Maintenance (PdM) that detects equipment anomalies in advance, ONEPREDICT is focusing on realizing 'Autonomous Factories (AI Native Factories)' where AI holistically analyzes and decides on data across equipment, processes, and quality for an entire plant.
"Government Must Also Lay Manufacturing AI Infrastructure"… Calls for Inter Ministerial Control Tower and SME Support
Yoon emphasized that for Korea to seize opportunities in the manufacturing AX race, the government's role is just as critical as technological development. Rather than competing head to head against nations deploying massive capital and infrastructure like the US and China, Korea needs strategic selection and concentration that leverages its manufacturing strengths.
"We must guard against spread thin government spending or turf wars between ministries, establishing a clear national roadmap around a unified control tower," he suggested. "Urgent priorities include AI reskilling for small and medium sized enterprises (SMEs), public AI infrastructure support centered around industrial complexes, and creating lighthouse factories to serve as exemplary cases."
Yoon believes that if manufacturing AI remains confined to large conglomerates, it cannot translate into industry wide national competitiveness. Support in data, infrastructure, and workforce development is essential so that SMEs can harness AI alongside major corporations.
In terms of technical differentiation, ONEPREDICT combines domain knowledge and physical laws with AI models beyond simple data analytics.
Yoon explained that by embedding domain knowledge and physical laws accumulated over 25 years in academia and industry into AI models, ONEPREDICT provides expert level solutions that present rationale, root causes, and actionable prescriptions.
Its core product lineup includes 'cyclone', a manufacturing data and MLOps platform, and 'pdx', an AI based integrated industrial asset management platform. Through its 'guardione' series predictive maintenance solutions that detect anomalies in rotating equipment and transformers in advance the company is achieving results such as reducing unexpected field downtime by over 30%.
"I take great pride in our technology that chemically embeds long researched field domain knowledge and physical laws into AI models, which is truly unique both domestically and globally," Yoon stated. "Rather than merely presenting results, we provide the reasoning and prescriptions together to assist expert level decision making."
He added, "Through 'technological assetization' that systematizes algorithms and workflows, we are simultaneously solving over 30 concurrent field projects. I feel the greatest fulfillment when our technology and products actually work on the shop floor and earn recognition from clients."
Yoon concluded, "We will continue to lead the global manufacturing AI market moving forward."
Source : Edaily (https://www.edaily.co.kr/News)
...
...
...
...
"Amid the dual crises of an aging workforce and population decline on the shop floor, AI Transformation is a matter of survival, not choice. Korea's global manufacturing competitiveness can only be safeguarded by advancing beyond simple process automation toward 'autonomous factories' where processes make decisions and prevent defects independently."
Byung-dong Yoon, CEO of ONEPREDICT, emphasized during an interview with E-Daily at the ONEPREDICT office in Banpo-dong, Seoul on the 7th that the future paradigm of manufacturing is shifting from "Automation (Smart Factory)" to "Autonomy (Manufacturing AX)."
As an active professor of Mechanical Engineering at Seoul National University who has long researched the integration of physical systems and AI, Yoon noted, "Past smart factories and today's manufacturing AX differ right from their core goals."
He explained: "While smart factories focused on 'automation' making processes run automatically the goal of manufacturing AX is 'autonomy,' where AI independently evaluates and takes corrective actions across quality inspection, maintenance, and process optimization. Rather than relying on human experience to inspect quality and resolve problems, the structure is evolving toward having AI agents embedded directly within processes to autonomously resolve issues."
<Source : Edaily>
Next After Smart Factory is 'Autonomy'… AI Evaluates Processes and Resolves Problems
Yoon analyzes that manufacturing AX has emerged as a central industry talking point due to the intersection of structural crises on the factory floor and the maturation of AI technology.
"As the working age population shrinks and veteran technicians retire, long accumulated tacit knowledge is vanishing," he said. "Just as urgency was mounting, data accumulated through smart factory initiatives combined with advancing AI technology, joining forces as a solution to this societal issue."
Yoon diagnosed that this situation could actually serve as a golden opportunity for South Korean manufacturing. While the US has already lost a significant portion of its shop floor master technicians, and Germany and Japan face delayed AI transformation due to slower Digital Transformation (DX) despite retaining veteran craftspeople, South Korea sits in a unique sweet spot.
Korea still retains experienced shop floor experts while having already progressed significantly in manufacturing DX.
Yoon evaluated: "Even though Germany and Japan have master technicians, they face a mismatch where digital transformation has lagged, resulting in a lack of data and delayed AI adoption. Korea still has experts with tacit knowledge on the ground, and because DX has progressed to accumulate data, we are in an advantageous position to rapidly embed this knowledge into AI for global competition."
In short, digitizing and embedding the experience and heuristics of skilled workers into AI before they leave the field will determine the competitiveness of Korean manufacturing.
"AI Finds Defect Causes That Humans Miss"… Beyond Predictive Maintenance to Autonomy
A prime domain requiring AI is mysterious quality issues that arise during manufacturing cases where components pass inspection individually, but unexplained defects occur after final assembly. Because multiple processes and countless variables are intricately intertwined, finding root causes through human intuition alone is nearly impossible.
Yoon compared the difference between humans and AI to the Go playing AI 'AlphaGo.'
"No matter how skilled a human expert is, human experience and vision have limits, making it impossible to calculate all complex variables," he stressed. "In contrast, AI can weigh all possibilities probabilistically to pinpoint and solve problems where tens of thousands of complex variables are entangled."
The goal of manufacturing AX does not stop at flagging anomalous signs. True autonomy means discovering anomalies, analyzing root causes, and directly orchestrating optimal responses.
Yoon is implementing the manufacturing AX paradigm he has long advocated on the front lines through ONEPREDICT, the industrial AI company he founded in 2016.
Beyond traditional Predictive Maintenance (PdM) that detects equipment anomalies in advance, ONEPREDICT is focusing on realizing 'Autonomous Factories (AI Native Factories)' where AI holistically analyzes and decides on data across equipment, processes, and quality for an entire plant.
"Government Must Also Lay Manufacturing AI Infrastructure"… Calls for Inter Ministerial Control Tower and SME Support
Yoon emphasized that for Korea to seize opportunities in the manufacturing AX race, the government's role is just as critical as technological development. Rather than competing head to head against nations deploying massive capital and infrastructure like the US and China, Korea needs strategic selection and concentration that leverages its manufacturing strengths.
"We must guard against spread thin government spending or turf wars between ministries, establishing a clear national roadmap around a unified control tower," he suggested. "Urgent priorities include AI reskilling for small and medium sized enterprises (SMEs), public AI infrastructure support centered around industrial complexes, and creating lighthouse factories to serve as exemplary cases."
Yoon believes that if manufacturing AI remains confined to large conglomerates, it cannot translate into industry wide national competitiveness. Support in data, infrastructure, and workforce development is essential so that SMEs can harness AI alongside major corporations.
In terms of technical differentiation, ONEPREDICT combines domain knowledge and physical laws with AI models beyond simple data analytics.
Yoon explained that by embedding domain knowledge and physical laws accumulated over 25 years in academia and industry into AI models, ONEPREDICT provides expert level solutions that present rationale, root causes, and actionable prescriptions.
Its core product lineup includes 'cyclone', a manufacturing data and MLOps platform, and 'pdx', an AI based integrated industrial asset management platform. Through its 'guardione' series predictive maintenance solutions that detect anomalies in rotating equipment and transformers in advance the company is achieving results such as reducing unexpected field downtime by over 30%.
"I take great pride in our technology that chemically embeds long researched field domain knowledge and physical laws into AI models, which is truly unique both domestically and globally," Yoon stated. "Rather than merely presenting results, we provide the reasoning and prescriptions together to assist expert level decision making."
He added, "Through 'technological assetization' that systematizes algorithms and workflows, we are simultaneously solving over 30 concurrent field projects. I feel the greatest fulfillment when our technology and products actually work on the shop floor and earn recognition from clients."
Yoon concluded, "We will continue to lead the global manufacturing AI market moving forward."
Source : Edaily (https://www.edaily.co.kr/News)