Asia\u2019s push into physical AI is being fueled by the same manufacturing prowess that made the region a global industrial powerhouse. Across South Korea, Japan, China and Taiwan, manufacturing remains a central pillar of economic growth. Unlike economies more heavily weighted toward services or software, these countries have long relied on large-scale production, export-driven industries, and highly optimized supply chains. That structural foundation is now shaping how artificial intelligence is adopted and where investment flows.<\/p>\n
Which makes it particularly significant that Config<\/a>, a Seoul- and San Jose- based startup building the data layer for robotic foundation models<\/a> (RFMs), has secured backing from the venture arms\u2019 of South Korea\u2019s biggest manufacturers.<\/p>\n Samsung Venture Investment led its oversubscribed $27 million seed round at a valuation of more than $200 million, bringing Config\u2019s total raised to $34 million. Hyundai Motor\u2019s venture arm ZER01NE Ventures, LG Tech Ventures, and SKT America, a South Korean telco giant\u2019s VC unit, also joined as strategic investors, alongside angel investor Pieter Abbeel (co-founder of Covariant AI and a UC Berkeley professor) and financial backers including Mirae Asset Ventures, Korea Development Bank, GS Futures, Kakao Ventures, and Z Ventures.<\/p>\n Config was founded in January 2025 by CEO Minjoon Seo, a former researcher at Meta and chief scientist at Twelve Labs, along with four co-founders with backgrounds at Waymo, Google and Naver. Instead of building robots themselves, the team is focused on a simpler goal, providing data robots need to learn and operate. They believe that better data will be key to making robots more useful.<\/p>\n Training large language models is expensive, because of the computing power required to process them, but the raw material, vast amounts of text from across the internet, is easy to obtain. Teaching robots to move is a completely different challenge, Seo said in an exclusive interview with TechCrunch. Every piece of training data has to be physically collected, like you need the robot, the facility to run it, and people to operate it. That makes robotics AI more costly to develop than software-only chatbot, according to the Seo. As companies pursue more capable robots, the cost of gathering and labeling data can rise quickly.<\/p>\n Config wants to be the company that makes everyone else\u2019s robot AI possible. The startup compares its role to TSMC, a Taiwanese chipmaker that manufactures \u00a0for Apple, Nvidia, and AMD without competing with any of them. Config aims to play a similar role in robotics by supplying the data. The approach is gaining traction as large manufacturers increasingly seek to build their own proprietary robot AI instead of relying entirely on outside vendors. That is the market Config is betting on. <\/p>\n Config is already generating revenue, COO and co-founder of Config Jack Bang said. The startup\u2019s current customers include large manufacturers, system integrators, and companies in the agriculture and defense sectors, Bang told TechCrunch. Peers in the space include Physical Intelligence, Generalist AI and Skild AI. <\/p>\n Techcrunch event<\/p>\n \n\t\t\t\t\t\t\t\t\tSan Francisco, CA<\/span> Config records humans performing physical tasks in controlled studio environments and in the field. The startup operates out of Seoul and Hanoi, where a workforce of nearly 300 handles data production. To date, it has accumulated over 100,000 hours of human motion data, more than 30 times the size of AgiBot World,<\/a> the largest comparable open-source dataset at roughly 3,000 hours<\/a>.<\/p>\n Most robotics teams train AI models on human motion data and then adapt those models for a robot. Config is taking a different approach, Seo said. The company focuses on transforming the data before training begins so it is better suited to the way robots move and interact with the world. Seo compared the process to language translation. Training a model on one type of data and expecting it to work seamlessly in another setting, Seo said, is trying to teach Korean using only English-language materials.<\/p>\n \u201cThe data must be converted, not the model. This conversion technology is Config\u2019s core technical differentiator,\u201d Seo said.<\/p>\n The funding will go toward three priorities: scaling its data operation in Vietnam and Seoul toward one million hours of collected data, growing its enterprise platform business to $10 million in ARR by the end of 2026, and launching a cloud-based Robot-as-a-Service product that lets companies run Config\u2019s foundation model without requiring onboard hardware.<\/p>\n<\/div>\n When you purchase through links in our articles, we may earn a small commission<\/a>. This doesn\u2019t affect our editorial independence.<\/em><\/p>\n
\n\t\t\t\t\t\t\t\t\t\t\t\t\t|<\/span>
\n\t\t\t\t\t\t\t\t\t\t\t\t\tOctober 13-15, 2026<\/span>\n\t\t\t\t\t\t\t<\/p>\n<\/p><\/div>\n<\/p><\/div>\n<\/div>\n![\"\"](\"https:\/\/techcrunch.com\/wp-content\/uploads\/2026\/05\/Config.jpg?w=680\")