[The Upcoming K-Robot Era]⑤"Korea, Strong in Manufacturing, Needs a Major Shift to an AI-Centric Robot Strategy"

Experts agree that for Korea to emerge as one of the world's top three powerhouses in the physical AI and humanoid robot industries alongside the United States and China, a fundamental shift toward an "AI-centric robot strategy" is essential. While some still approach the robot industry from the perspective of precision machinery, components, and manufacturing technology, others emphasize that a paradigm shift is required to view it as part of the AI industry rather than the manufacturing sector.

Kim Ikjae, Director of the AI & Robotics Research Division at the Korea Institute of Science and Technology (KIST), said on March 10, "Robots are no longer just 'precise machines,' but rather 'mediums through which intelligence acts in the physical world,'" adding, "The future competitiveness of robots will depend not on how precisely they are made, but on how intelligently they can act." Kim emphasized, "Ultimately, for Korea to become a robotics powerhouse, the most important starting point is to move away from a manufacturing-centric mindset and shift to an AI-driven industrial strategy."

In particular, in order for Korea to differentiate itself between the software strengths of the United States and the hardware competitiveness of China, experts believe Korea must focus on developing robots specialized for industrial applications.

An Sungwon, Head of the AI Policy Research Department at the Software Policy & Research Institute, stated, "For physical AI, represented by robots and autonomous driving technology, to take root across Korean industries, there is a need for 'domain-specialized high-performance AI models' from the software side." An explained, "Domain-specialized high-performance AI models learn data from specific industrial sites and synchronize the 'perception-inference-execution' process in real time to adapt and revise responses according to the situation," adding, "Since the key to competitiveness in physical AI is securing high-quality, refined data, Korea should leverage its strengths rather than directly competing with the U.S. and Chinese big tech companies with more advanced technology and capital."

Son Dongshin, a member of the Future Robotics Lab at LG CNS, commented, "Hands are crucial for field operations, and human tasks require precision and delicacy, so the core lies in manipulation technology." He added, "Even China still faces hardware challenges when it comes to manipulation technology that utilizes hands."

When it comes to fourth-generation robots capable of precision work, there are still areas where the United States has yet to achieve commercialization. In terms of robots using hands and arms-especially fingers-for actual tasks, Korea holds a competitive edge, and this could be an opportunity for Korea to take the lead in application domains by creating real-world use cases. Son pointed out, "As Korea is a manufacturing powerhouse, there are many places where robots can be applied," adding, "We need to quickly create cases in factories, logistics centers, and other industrial sites, and, based on these, develop proven solutions and rapidly advance in the application domain."

원본보기 아이콘

Experts suggest that, since the country that sets the standards secures leadership in the robotics industry, the Korean government should take a proactive role in global discussions and expand joint research efforts.

Kim stated, "In the robotics industry, standards equate to the market because robots are not standalone products but integrated systems combining hardware, software, AI, and data." He continued, "Key standard areas going forward could include robot behavior expression frameworks, environmental perception data structures, and AI-robot interfaces. If Korea can secure structural standards, various manufacturers will develop on top of them, services will interconnect, data will accumulate, and an industrial ecosystem will form. The government should serve not just as a participant, but as a proposer of standards." He explained that this requires efforts such as global joint research, participation in test beds, and contributions to open ecosystems.

Data standardization is also an urgent issue. An pointed out, "Since each company has different ways and volumes of collecting data, data standardization (integration) is not easy in practice," stressing, "The government should actively take the lead in establishing standards." Referring to past attempts at data standardization at Seoul National University Bundang Hospital and Ajou University Hospital, An noted, "Each hospital had different data formats and conflicting interests, making sharing difficult." He added, "If accumulated data can be shared, it can reduce the trial and error experienced so far. This would help small and medium-sized enterprises and startups, which often struggle to secure data," and said, "There are ongoing discussions in the field about standardization that would allow a single program to control movement, regardless of the number of robot joints."

There are also calls for establishing systems that can be quickly and safely applied to humanoids, just as standards and certification systems were previously established for autonomous delivery robots. Son said, "It is necessary to expand regulatory sandboxes and pilot environments and develop models that allow demand-side companies to rapidly adopt and validate domestic robot companies." He also noted that humanoids, which include hardware, pose the challenge of high verification, testing, and pilot costs, but if the government expands the environment through programs, regulatory special zones, and regulatory sandboxes, the domestic robot industry will be able to take advantage of these opportunities.

Experts anticipate that the robot industry could offer solutions to structural challenges such as population decline and skilled labor shortages. Son said, "The adoption of robots is not simply about reducing labor costs. It enables a reorganization of the domestic production system centered on highly skilled robot operators," adding, "This is a strategy that simultaneously addresses the structural challenge of population decline and maintains or strengthens industrial competitiveness, making 'reshoring strategies' to bring factories previously moved overseas back to Korea a viable long-term scenario."

On this topic, Song Changjong, Head of the Physical AI Team at the Ministry of Science and ICT, said, "Physical AI is not something that a single government-funded research institute or a single company can handle alone; it requires a highly agile organization." He continued, "If private organizations, the government, and research institutes do not work together, we will not be able to compete with the U.S. and China, so it is crucial for companies and the government to form a united team, and funding programs are being focused in that direction."

Related Articles: [The Upcoming K-Robot Era]

①China, Once Obsessed with Building the Shell, Now Focuses on Brains... Hardware to Software Competitiveness Shift

②MWC Becomes a Battleground for Chinese Robots... From Hardware to Robotics AI

③Korea Caught Between U.S.-China Rivalry... Lacking a Full-Stack Ecosystem

④Korea's Edge Is 'Data'... Differentiation in Software Possible Through Accumulation and Utilization

© The Asia Business Daily(www.asiae.co.kr). All rights reserved.