South Korea Has a Robotics Problem: Rainbow Robotics, Hyundai, Samsung, and the Third Front Nobody Is Watching

South Korea has the highest robot density of any country on Earth. For every 10,000 manufacturing workers, there are over 1,000 industrial robots. The next closest country, Singapore, has roughly half that number. Germany has about a third. The United States has about a quarter.

South Korea is also the country where Hyundai Motor Group spent $1.1 billion to acquire Boston Dynamics. Where Samsung Electronics made a strategic investment in Rainbow Robotics. Where a university lab built a humanoid robot that won the hardest robotics competition ever staged. Where the government has a national robot industry strategy backed by law.

And South Korea has shipped 10 humanoid units.

Ten. Total. From one company. Rainbow Robotics, a publicly traded KAIST spin-off in Daejeon with 100 to 250 employees, listed on the KOSDAQ under ticker 277810.KQ.

Until this article, the robotsin.life tracker had zero articles about Rainbow Robotics or South Korea’s humanoid ambitions. That is a gap in our coverage, and it reflects a broader blind spot in how the humanoid race is discussed. The conversation focuses almost entirely on the US-China dynamic. But South Korea sits in between those two powers - geographically, economically, and technologically - and the pieces it holds could reshape the race.

The HUBO lineage

The Rainbow Robotics story starts in a lab at KAIST, the Korea Advanced Institute of Science and Technology, in Daejeon. KAIST is often described as South Korea’s MIT, and for robotics, the comparison is not an exaggeration.

Professor Oh Jun-ho led KAIST’s humanoid robotics laboratory for decades. In 2004, his team built HUBO, one of the first full-size humanoid robots in Asia. HUBO could walk, grasp objects, and perform basic manipulation tasks. It was impressive for its time, but it was the kind of achievement that gets a mention in IEEE Spectrum and then fades from public memory.

What happened in June 2015 changed that.

DARPA, the US Defense Advanced Research Projects Agency, organized the DARPA Robotics Challenge (DRC) Finals in Pomona, California. The DRC was, by a wide margin, the most demanding robotics competition ever held. Teams had to build humanoid robots that could drive a vehicle, open a door, turn a valve, drill through a wall, cross rubble, and climb stairs. All autonomously, with degraded communications, in under one hour.

Twenty-three teams from around the world competed. The winner was Team KAIST with DRC-HUBO, scoring 8 out of 8 tasks in 44 minutes and 28 seconds. They beat NASA’s Jet Propulsion Laboratory team. They beat MIT. They beat Carnegie Mellon. They beat every American, Japanese, and European team in the field.

DRC-HUBO won for a specific, clever reason. While most competing robots were fully bipedal and walked upright through every task, DRC-HUBO had a transforming design. It could stand and walk on two legs, but it could also kneel and drive on wheels mounted on its knees. This gave it stability during tasks that required precise manipulation (no risk of falling over) and speed during transit between tasks. It was an engineering decision that prioritized completing the mission over looking impressive.

Professor Oh Jun-ho took the expertise from two decades of humanoid research at KAIST and founded Rainbow Robotics in 2011. The company was, from its inception, meant to be the commercial vehicle for KAIST’s robotics technology.

Rainbow Robotics today

Rainbow Robotics operates from Daejeon, near the KAIST campus, with 100 to 250 employees. The company is publicly traded on KOSDAQ, giving it access to public market capital that most robotics startups do not have. Its stock symbol, 277810.KQ, is actively traded, and the company’s market capitalization has fluctuated significantly as humanoid robotics hype has ebbed and flowed.

The company builds two main product categories.

Collaborative robot arms (RB-series). Rainbow’s bread-and-butter business is its line of collaborative robot arms, which compete with Denmark’s Universal Robots, the global market leader in cobots. The RB-series includes 3 kg, 5 kg, 10 kg, and 16 kg payload variants. These are proven commercial products with real revenue, real customers, and real deployments across Korean manufacturing. This business generates the revenue that keeps Rainbow alive while the humanoid program develops.

RB-Y1 humanoid mobile manipulator. Rainbow’s humanoid platform is the RB-Y1, a mobile manipulator with dual dexterous arms mounted on a wheeled base. It is not a bipedal humanoid in the traditional sense. It does not walk on two legs. But it has a humanoid upper body with arms and hands capable of sophisticated manipulation, and it is designed for the same commercial applications - manufacturing assistance, logistics, service - that bipedal humanoids target.

The decision to build a wheeled mobile manipulator rather than a bipedal walker is pragmatic. Wheels are more reliable than legs. They are cheaper to engineer. They do not fall over. For the specific industrial applications Rainbow is targeting - tasks in Samsung factories, Korean electronics manufacturing, automotive assembly - a wheeled base with a humanoid upper body may actually be more useful than a walking platform.

But it also means Rainbow is playing a slightly different game than the companies making headlines. When people say “humanoid robot,” they picture something that walks on two legs. Rainbow’s HUBO walked on two legs and won the DARPA Challenge doing it. The RB-Y1 does not. The disconnect between Rainbow’s legacy and its current product is one reason the company gets less attention than its credentials deserve.

The Samsung connection

In 2024, Samsung Electronics made a strategic investment in Rainbow Robotics. The details of the investment amount were not fully disclosed, but Samsung’s involvement goes beyond writing a check.

Samsung Electronics is one of the largest manufacturers in the world. Its semiconductor fabrication plants, smartphone assembly lines, and display panel factories are among the most automated and technologically sophisticated manufacturing environments anywhere. Samsung also operates Samsung Research, one of the biggest corporate R&D organizations globally, with active programs in AI, robotics, and automation.

Samsung’s investment in Rainbow Robotics mirrors a pattern seen elsewhere in the humanoid race. BYD invested in AgiBot. BMW partnered with Figure AI. Mercedes-Benz partnered with Apptronik. In each case, a major manufacturer provides both capital and, more importantly, a real-world deployment environment for humanoid robots. The manufacturer gets early access to automation technology. The robotics company gets paying customers, operational data, and credibility.

For Rainbow, the Samsung partnership is transformative in potential if not yet in scale. Samsung’s factories represent a massive deployment opportunity. If Rainbow’s RB-Y1 can prove itself in Samsung’s semiconductor or electronics manufacturing lines, the pathway to thousands of units becomes visible. Samsung’s global manufacturing footprint spans South Korea, Vietnam, India, and beyond. A successful pilot could scale across continents.

The problem is that “potential” and “deployed” are different words. As of early 2026, Rainbow has shipped approximately 10 humanoid units. Samsung’s investment has not yet translated into large-scale deployments. The partnership is real, but the production numbers suggest it remains in its early stages.

Hyundai’s dual position

South Korea’s position in the humanoid race gets even more interesting when you account for Hyundai Motor Group.

In June 2021, Hyundai completed its acquisition of Boston Dynamics for approximately $1.1 billion. Hyundai acquired an 80% stake, with SoftBank retaining 20%. This made Hyundai the owner of arguably the most famous robotics company in the world - the company that built BigDog, Spot, and Atlas.

Boston Dynamics has shipped approximately 1,000 humanoid units as of early 2026, placing it among the top five companies globally. Its Electric Atlas, unveiled in production-ready form at CES 2026, represents one of the most technically advanced humanoid platforms ever built. The company also announced an AI partnership with Google DeepMind, integrating Gemini Robotics foundation models with Atlas.

But here is the nuance that makes South Korea’s situation uniquely complex. Hyundai owns Boston Dynamics, but Boston Dynamics operates from Waltham, Massachusetts. Its engineers are American. Its manufacturing will be American. Its technology development happens in America. Hyundai’s ownership means Korean capital funds the company, and Hyundai’s strategic direction influences its roadmap. But the robots, the talent, and the production are firmly American.

Meanwhile, Hyundai Motor Group has its own robotics ambitions beyond Boston Dynamics. Hyundai has been developing factory automation robots, logistics robots, and wearable exoskeletons through its internal robotics division. The company has stated its goal of becoming a “smart mobility solution provider,” with robotics as a core pillar alongside electric vehicles and autonomous driving.

Hyundai also operates Hyundai Robotics (a subsidiary focused on industrial robot arms) and has partnered with various Korean AI and robotics startups. The conglomerate’s factory network - spanning automobile assembly plants in South Korea, the United States, Czech Republic, Turkey, India, Brazil, Indonesia, and China - represents one of the largest potential deployment environments for humanoid robots in the world.

The question is whether Hyundai will deploy Boston Dynamics robots in its Korean factories, develop its own humanoid platform, or partner with Korean companies like Rainbow Robotics for domestic deployment. All three paths have been discussed in Korean industry media. None has produced visible results at scale yet.

The world’s most automated country

To understand why South Korea matters in the humanoid race, you have to understand how deeply automation is already embedded in the Korean economy.

South Korea’s robot density of over 1,000 industrial robots per 10,000 manufacturing workers is not a statistical fluke. It reflects decades of deliberate industrial policy. The Korean government passed the Intelligent Robot Development and Promotion Act in 2008, creating legal and institutional frameworks for robotics development. The Robot Industry Promotion Act followed, establishing dedicated funding mechanisms and R&D programs.

Korean manufacturers adopted automation faster and more thoroughly than their counterparts in any other country. Samsung, Hyundai, LG, SK, and other Korean conglomerates (chaebols) invested heavily in robotic automation for electronics assembly, automobile manufacturing, semiconductor fabrication, and shipbuilding. The result is a manufacturing sector that already operates with more robots per worker than anywhere else on Earth.

This creates a paradox for humanoid robotics. On one hand, Korea has the world’s most automation-ready industrial culture. Korean manufacturers understand robots. They have procurement processes for robots. They have engineering teams that integrate robots. They have factory floors designed to accommodate robots. If any country should be a natural market for humanoid robots, it is South Korea.

On the other hand, Korea’s existing robot fleet is overwhelmingly traditional industrial robots - fixed-base arms from companies like FANUC, ABB, KUKA, and Hyundai Robotics itself. These robots are excellent at the repetitive, high-precision tasks that dominate Korean manufacturing. The business case for replacing them with more expensive, less reliable humanoid robots is not yet clear.

The K-robotics ecosystem

South Korea’s robotics ecosystem extends well beyond Rainbow Robotics and Hyundai’s investments. The country has a deep bench of capabilities that could become relevant as the humanoid race evolves.

KAIST and Korean universities. KAIST’s humanoid robotics research is world-class, as the DARPA Challenge victory demonstrated. But KAIST is not alone. Seoul National University, POSTECH, Yonsei University, and Korea University all have active robotics programs. Korean universities produce thousands of robotics-capable engineers annually. The talent pipeline is strong.

Korean AI capabilities. South Korea is a serious player in AI research, with Samsung AI Center, Naver (Korea’s equivalent of Google), and Kakao all investing in AI development. Naver’s CLOVA AI platform, Samsung’s on-device AI for Galaxy devices, and various Korean AI startups provide a foundation for the AI systems that humanoid robots require.

The semiconductor advantage. Samsung is the world’s largest semiconductor manufacturer by revenue (alternating with Intel depending on the year), and SK Hynix is the second-largest memory chip maker. South Korea’s semiconductor expertise is directly relevant to humanoid robotics because every humanoid robot needs processors, memory, sensors, and custom chips. Korean companies can potentially develop specialized AI inference chips optimized for robotics applications.

Manufacturing culture. Korean manufacturing culture emphasizes continuous improvement (a legacy of Japanese lean manufacturing principles adapted for Korean contexts), rapid iteration, and extreme attention to quality. This culture is well-suited to humanoid robot production, where hardware reliability is critical.

Why Korea has not shipped more

Given all of these advantages, why has South Korea shipped only 10 humanoid units? The answer involves several intersecting factors.

The chaebol decision cycle. Korean conglomerates are powerful but deliberate. Hyundai’s acquisition of Boston Dynamics was a $1.1 billion bet on robotics, but translating ownership into Korean-manufactured humanoid robots is a multi-year strategic process involving multiple business divisions, board approvals, and integration plans. Samsung’s investment in Rainbow Robotics similarly reflects a strategic interest that has not yet matured into large-scale deployment.

Korean chaebols think in decades, not quarters. This is a strength for long-term technology development but a weakness in a race where Chinese startups go from founding to shipping thousands of units in under two years.

The US-China squeeze. South Korea sits between its security alliance with the United States and its economic interdependence with China. This creates complications for robotics strategy. Should Korean companies develop humanoid robots using American AI (from OpenAI, Google DeepMind, or NVIDIA) or develop indigenous alternatives? Should they target Chinese or American markets, knowing that geopolitical tensions could disrupt either? Should Hyundai deploy Boston Dynamics robots or develop Korean-made alternatives?

These questions do not have easy answers, and the ambiguity slows decision-making.

The domestic market question. South Korea’s manufacturing sector is large but not as large as China’s or America’s. The total domestic market for humanoid robots may be smaller than what Chinese companies can access through their home market alone. Korean companies may need to compete internationally from day one, which requires a different (and harder) go-to-market strategy than serving domestic customers first.

Risk culture. Korean corporate culture, particularly in the chaebols, is risk-averse relative to Silicon Valley startups or Chinese tech companies. Major investments require extensive study, consensus-building, and senior leadership buy-in. The speed with which Brett Adcock raised $1.85 billion for Figure AI or Peng Zhihui went from founding to factory at AgiBot would be difficult to replicate within the Korean corporate system.

The timeline of Korean humanoid robotics

What the third front looks like

The humanoid race is typically described as a two-front contest between the United States and China. The numbers support this framing. China’s six tracked companies have shipped over 13,000 units. America’s five companies have shipped over 2,000. Everyone else combined has shipped fewer than 100.

But framing the race as purely US vs. China misses an important dynamic. South Korea is not just a spectator. It is a player with assets that neither the US nor China can easily replicate.

Consider: South Korea owns the company (Boston Dynamics) with the most recognizable humanoid robot brand in the world. It has a publicly traded humanoid manufacturer (Rainbow Robotics) backed by one of the world’s largest electronics companies (Samsung). It has the highest robot density on the planet, meaning its workforce and factory infrastructure are already adapted to working alongside robots. It has world-class semiconductor manufacturing that could produce custom AI chips for robotics. And it has a DARPA Challenge-winning heritage that proves Korean engineers can build humanoids that outperform the best in the world.

The pieces are all there. The question is whether they will come together.

There are three scenarios for how South Korea’s role in the humanoid race could evolve.

Scenario 1: The integrator. Hyundai deploys Boston Dynamics Atlas robots in its Korean and global factories while simultaneously supporting Rainbow Robotics’ expansion through Samsung partnerships. Korean semiconductor companies develop custom AI chips for humanoid applications. South Korea becomes the first country to deeply integrate humanoid robots into an already heavily automated manufacturing sector, creating a deployment model that other countries follow. This is the best-case scenario, and it requires active coordination between chaebols that traditionally compete with each other.

Scenario 2: The acquirer. Korean companies continue to acquire foreign robotics capabilities rather than building domestic ones. Hyundai already owns Boston Dynamics. Samsung could acquire additional robotics companies. South Korea becomes a financial player in the humanoid race, funding American and potentially European robotics companies while Korean domestic manufacturing remains dependent on foreign-designed humanoids. This is comfortable for the chaebols but leaves Korea strategically dependent.

Scenario 3: The bystander. Decision-making takes too long. The chaebols study the market while Chinese and American companies ship tens of thousands of units. Rainbow Robotics remains a small player with impressive credentials but limited market impact. Hyundai’s ownership of Boston Dynamics provides financial returns but does not translate into Korean manufacturing capability. South Korea’s robotics talent continues to emigrate. By the time Korean companies are ready to move, the market positions are locked in. This is the scenario that Korea’s current pace of 10 shipped units suggests.

The numbers in context

Where does South Korea actually stand in the global humanoid race?

If you count Boston Dynamics’ 1,000 units as Korean (because Hyundai owns the company), the picture changes dramatically. Korea-affiliated humanoid shipments would jump to 1,010, making it the third largest player behind China and the rest of the US. But attributing American-manufactured, American-engineered robots to Korea because a Korean conglomerate owns the parent company is a stretch that most analysts would not make.

The more honest assessment is that South Korea’s domestic humanoid manufacturing capability is represented by Rainbow Robotics and its 10 units. That is the reality on the ground. Everything else - Hyundai’s ownership of Boston Dynamics, Samsung’s investment in Rainbow, the KAIST heritage, the semiconductor advantage - is potential. Potential matters, but it does not ship robots.

What Rainbow needs

Rainbow Robotics specifically needs three things to become a meaningful player in the humanoid race.

Scale deployment through Samsung. The Samsung investment needs to translate into hundreds, then thousands, of RB-Y1 units deployed in Samsung factories. If Samsung commits to replacing specific manual tasks in its semiconductor fabs and smartphone assembly lines with Rainbow robots, the production volumes follow. Samsung makes this decision based on demonstrated ROI in pilot programs. Those pilots need to succeed.

A bipedal platform. The RB-Y1 is a practical machine, but the market increasingly views bipedal locomotion as a requirement for true humanoid robots. Rainbow has the HUBO heritage. The engineering knowledge to build walking robots exists in the company’s DNA. A bipedal platform would position Rainbow as a full-spectrum humanoid manufacturer and attract attention from customers and investors who currently overlook wheeled platforms.

Speed. Rainbow needs to compress timelines. The company was founded in 2011. It has shipped 10 humanoid units in 15 years. AgiBot was founded in 2023 and shipped 5,200 units in under three years. The pace differential is enormous. Whether the acceleration comes from Samsung increasing its commitment, the Korean government providing manufacturing subsidies, or Rainbow’s own management deciding to move faster, something has to change.

The lesson of the third front

South Korea’s position in the humanoid race teaches something about the difference between having the right components and having a working system.

Korea has world-class robotics research. It has the world’s most automation-ready manufacturing culture. It has two massive conglomerates investing in humanoid robotics. It has semiconductor manufacturing that could produce custom robotics chips. It has a publicly traded humanoid manufacturer with a DARPA Challenge-winning heritage.

It has shipped 10 humanoid units.

The components do not self-assemble. They need a catalyst - a decision, a policy, a partnership, a competitive shock - that turns latent capability into active production. China had that catalyst in the form of MIIT guidelines, provincial subsidies, and a manufacturing ecosystem primed for new hardware categories. The United States had it in the form of Silicon Valley venture capital and visionary (or at least loudly ambitious) founders. South Korea has not found its catalyst yet.

But dismissing Korea would be a mistake. The country has a documented history of moving from seemingly peripheral positions to global leadership in remarkably short timeframes. Korea went from making cheap textiles in the 1960s to producing the world’s best semiconductors and smartphones within two generations. The pattern - government direction, chaebol execution, relentless focus on manufacturing excellence - has worked before in multiple industries.

If Hyundai and Samsung decide, truly decide, to make humanoid robotics a strategic priority on the same level as semiconductors, EVs, or smartphones, South Korea could compress a decade of catch-up into three or four years. The infrastructure, the talent, the capital, and the manufacturing culture are all in place.

The third front is quiet. But quiet is not the same as empty.