In a major cross-border funding move in the robotics AI space this year, Physical AI startup RLWRLD has announced a $15 million seed round backed by a strategic coalition of industrial titans from Korea and Japan. The round includes LG Electronics, SK Telecom, KDDI, ANA Holdings, Mitsui Chemicals, and Shimadzu Corporation, alongside top-tier VC firms Hashed, Mirae Asset, and Global Brain.
RLWRLD is building foundation models for robotics trained on real-world sensor data, industrial workflows, and dynamic environments. While generative AI companies are racing to perfect text, image, and code models, RLWRLD is focused on enabling machines to interact intelligently with the physical world, a field often described as Physical AI.
The $15 million in funding will primarily be used to deploy pilot projects with strategic investors across Korea, Japan, and India; build foundational AI models trained on vast real-world datasets; develop robotic dexterity, specifically five-finger manipulation skills; expand computing infrastructure with high-performance GPUs; and recruit world-class AI and robotics talent.
Powering the future of physical AI
Founded by serial entrepreneur Jung-hee Ryu in 2024 in South Korea, RLWRLD’s core technology combines large language models (LLMs) with robotics control systems to enable industrial robots to reason, adapt, and act autonomously. These AI-powered robots will be capable of executing complex tasks in factories, warehouses, logistics hubs, and service environments, going far beyond traditional industrial machines’ repetitive, pre-programmed motions.
According to the International Federation of Robotics (ifr), more than 540,000 new industrial robots were deployed globally in 2023, bringing the total number of active units to over 4 million. However, most of these robots are rigid, brittle systems that struggle with tasks requiring dexterity, adaptability, or decision-making.
RLWRLD aims to bridge this gap by creating intelligent agents that learn from real-world data, rather than relying solely on simulation or hard-coded rules. The company’s roadmap includes proof-of-concept projects with its strategic investors, acquiring robotic hardware, expanding GPU infrastructure, and hiring elite AI researchers and roboticists to push the field forward.
The competition: RLWRLD vs. Figure AI and Agility Robotics?
The Korean startup is also exploring advanced five-finger hand movements, a level of robotic dexterity not yet demonstrated by leading competitors such as Figure AI and Agility Robotics. This leap could unlock possibilities in precision assembly, food service, and elderly care.
Agility Robotics: The Oregon-based startup has carved out a strong position in the humanoid robot race with its flagship robot, Digit. Designed to work alongside humans in warehouses and logistics centers, Digit is already being piloted by major retailers like Amazon. The company was recently in the headlines as it eyes to raise $400 million at a $1.75 billion valuation and is scaling up manufacturing. However, Agility’s core focus remains on form, building human-like mobility and hardware while leaving the intelligence stack relatively limited in complexity.
Figure AI: Backed by heavyweights like OpenAI, Microsoft, and NVIDIA, Figure AI has become the poster child for humanoid robotics with its vision of building robots that look, move, and work like humans. Its $1.5 billion funding round valuing the company at nearly $40 billion and lofty goal of producing 100,000 humanoid robots signal serious intent. Figure’s humanoids are being developed for tasks in manufacturing and general labor, but much of their autonomy still depends on scripted behaviors or third-party AI modules.
While Figure AI and Agility Robotics focus heavily on the humanoid form factor, RLWRLD is building the brain to drive a wide array of robot forms, from articulated arms to mobile units.
What fuels RLWRLD?
By uniting AI model innovation with real-world industrial expertise, RLWRLD aims to create intelligent robotic systems that can work safely and productively alongside humans in complex environments. Its technology could redefine how industries think about robotics as tools and collaborative, adaptive co-workers. RLWRLD’s real-world-first approach and cross-border backing could give it a unique path to impact.
“Silicon Valley is building brains for the internet. We’re building brains for machines,” said Jung-hee Ryu, Founder & CEO, RLWRLD.
“We’re not just writing papers—we’re training models that solve real-world problems on real production floors,” added Jinwoo Shin, Chief Scientist, RLWRLD & KAIST Chair Professor.
“RLWRLD combines deep AI research with industrial execution at scale. That’s what makes this a rare, high-conviction investment,” added Simon Seojoon Kim, CEO, Hashed.
