A viral video of humanoid robots dancing during Chinese New Year celebrations captured the world’s imagination. It was more than a spectacle. It showed how quickly humanoid robotics is moving from research labs and industrial settings into public life, cultural moments, and everyday environments.
The performance also raised a bigger question: Who is truly shaping the future of humanoid robotics, and how can innovation strength be measured beyond headlines, demos, and media attention?
That question matters because humanoid robotics is no longer only an engineering challenge. It is becoming a strategic technology field, supported by growing government attention, expanding commercial expectations, and rapid advances across robotics, AI, and automation.
China has become one of the most active policy drivers, moving from innovation support toward standardization, testing, and real-world deployment. Recent government activity includes dedicated standardization work for humanoid robots and embodied intelligence, as well as initiatives focused on validating robots in industrial, service, and special-purpose environments.
China is not the only country putting humanoid robotics on the strategic agenda. South Korea has launched the K-Humanoid Alliance, a government-backed initiative that brings together industry, academia, and research organizations with the ambition of becoming a global leader in humanoid robots by 2030. Other countries, including Japan and the United Kingdom, are also addressing humanoids or adjacent robotics fields through broader robotics strategies, technology assessments, and deployment programs.
The humanoid robotics market is still emerging, but expectations for future growth are substantial. Short-term forecasts from Grand View and Goldman Sachs estimate an approximately USD 40B market by the early to mid-2030s. Looking further ahead, Morgan Stanley suggests the humanoids market could surpass USD 5T by 2050, including related supply chains, repair, maintenance, and support.
While consumer and home use cases remain part of the long-term vision, analysts expect growth to start in structured industrial environments, where labor shortages, repetitive or physically demanding tasks, and the need for flexible automation create a clear business case.
Because patent information can serve as a forward-looking indicator of future commercial applications and indirect business potential, it offers a valuable lens for understanding who may shape this emerging market. By looking at patent activity across the core technologies that enable humanoid robots, it becomes possible to identify who is investing, where innovation is concentrated, and which companies may be building defensible positions in the field.
Patent analytics is especially powerful when a field can be described through clearly defined technologies. In application-driven fields such as humanoid robotics, the boundaries are more complex. Patent documents often describe the technical invention in detail, but they do not always clearly state the final application context. A patent for a gripper, tactile sensor, motion control method, or planning architecture may be highly relevant to humanoid robotics, but the patent itself may also apply to industrial robots, medical devices, automated logistics systems, or other machine types. As a result, no single patent query or technology class can perfectly capture the humanoid robotics field.
However, this is also where patent analytics adds significant value. By combining carefully designed technology classifiers with ownership, portfolio size, portfolio strength, and concentration metrics, it becomes possible to narrow the field and distinguish broad robotics patent holders from companies with a more focused humanoid robotics profile. The analysis may not capture every application-specific nuance, but it helps move the discussion from market visibility and public demonstrations to measurable innovation activity.
Humanoid robotics cannot be captured by one patent class or keyword set alone. The field spans multiple technical domains, so we focused on three core technology areas that make humanoid robots functional, adaptive, and autonomous:
This approach helps surface the companies with relevant patent holdings across the technologies that make humanoid robots functional, adaptive, and autonomous.
Our search was designed to capture the core technologies that make humanoid robots physically functional: interaction systems, morphology, and control and planning architectures. However, humanoid robotics also depends on several enabling technologies that may sit outside this framework, including AI learning models, perception systems, batteries and power management, compact actuators, safety systems, manufacturing processes, and fleet-level software infrastructure. These areas are critical for commercialization, but they are often patented as general robotics, AI, automation, or component technologies rather than as humanoid robotics inventions. As a result, the analysis focuses on the technologies most directly tied to humanoid embodiment and operation, while recognizing that the broader humanoid robotics ecosystem extends beyond our search.
The analysis includes more than 26,000 patent families across a wide span of technology clusters.
We started the analysis by looking at how the three core technologies developed over time.
Interaction systems and control and planning architectures are essential to humanoid robotics, but they are not exclusive to it. These technologies also appear across industrial robotics, collaborative robots, warehouse automation, service robotics, medical robotics, autonomous mobility, drones, agricultural robotics, and other fields where robots must manipulate objects, coordinate movement, or operate autonomously. Because they are relevant to such a broad range of applications, patent filings in these areas began increasing in the mid-2010s and have shown steady growth since then.
Morphology provides a more humanoid-specific lens because it captures patents tied to human-like body structures, bipedal movement, articulated limbs, and whole-body coordination. This makes the recent trend especially notable: after nearly a decade of relatively flat activity, morphology-related patent filings began to rise sharply in 2023, suggesting growing innovation around the physical embodiment of humanoid robots.
Then we switched the lens and analyzed where patents related to the three key humanoid robotics technologies were invented. This inventor-level view helps reveal where the underlying research and development activity is taking place, regardless of where the patent owner is headquartered.
The results show a clear shift toward China. Similar to other modern technology fields, Chinese inventors are becoming increasingly prominent across the humanoid robotics landscape. Over the past decade, patent activity from China-based inventors has grown by approximately 2x in interaction systems, 2.5x in morphology, and 5x in control and planning architectures.
Because China has been known for high levels of patent filing activity over the past decade, we also examined the increase in portfolio strength, as measured by the LexisNexis® Patent Asset Index, to better understand the impact of Chinese inventions. Patent Asset Index, a metric designed to measure innovative strength rather than portfolio size alone.
The value of patents varies significantly depending on their scope and utility. While some patents encompass foundational technologies applicable across multiple applications, others may cover incremental enhancements to niche features with limited practical use. Consequently, simply counting patent families may not accurately reflect the true value or enforceability of a patent portfolio. A reliable indicator of a patent family’s value is the extent to which other innovations build upon the technology it protects (Technology Relevance), and the breadth of protection (Market Coverage) deemed appropriate by its ultimate owner. The Patent Asset Index serves as a comprehensive measure of the innovative strength of a patent portfolio. This metric aggregates the Competitive Impact (average quality) scores of individual patent families within the portfolio, providing a relative value benchmark against other patents in the same technological domain.
When looking specifically at morphology-related patents, China leads by active portfolio size, at 73% by the end of 2025, and also holds the strongest overall portfolio, at 63% by the end of 2025, even if its lead in portfolio strength is less pronounced than its lead in volume. But the ranking changes for other authorities when switching from quantity to quality. In 2025, South Korea ranks second by portfolio size with 11%, followed by Japan with 7% and the United States with 5%. When looking at portfolio strength through the LexisNexis® Patent Asset Index, however, the United States moves into second place with 11%, ahead of Japan with 10% and South Korea with 7%. This shows why portfolio size alone does not fully reflect the strength of a country’s patent position. A smaller portfolio can carry greater weight when its patents show higher technological relevance and broader market coverage.
The pattern remains consistent when looking at portfolios across the three aggregated technology fields.
Each bubble represents the active patent portfolio for a given inventor location at the end of a reporting year. The x-axis shows portfolio size, while the y-axis shows Competitive Impact, indicating the average strength of the patent portfolio. Bubble size represents portfolio strength as measured by the LexisNexis® Patent Asset Index. The color intensity shows the time progression: the lightest bubble represents the portfolio as of Dec. 31, 2020, and the darkest bubble represents the portfolio as of Dec. 31, 2025.
China’s portfolio size and Patent Asset Index show sustained acceleration from 2020 to 2025. This indicates strong R&D activity and active patent filing strategies in humanoid robotics. This development aligns with China’s broader robotics push during the 14th Five-Year Plan period, launched in 2021, when the government identified robotics as a strategic industry and prioritized core technologies highly relevant to humanoid robots, including bionic perception, intelligent control, integrated joints, electronic skin, and natural human-robot interaction.
The United States consistently achieves the highest Competitive Impact, indicating that its humanoid robotics patent portfolio is smaller than China’s but, on average, stronger in terms of technological relevance and market coverage.
Japan maintains steady growth in both portfolio size and Patent Asset Index, but its Competitive Impact gradually declines after 2021.
Germany and South Korea show more modest portfolio growth, but both display an increase in Competitive Impact in the most recent years.
We also looked at how patents are distributed across different owner types. The results show that companies account for the largest share of humanoid robotics patent portfolios in all five major inventor locations, but the balance differs significantly by country. Japan, Germany, and the United States are strongly company-led, with companies representing 92%, 83%, and 86% of portfolio size, respectively. China shows a more mixed structure, with companies accounting for 58% and research institutions representing a substantial 39%. South Korea stands out with the most diversified ownership profile: companies account for 45%, research institutions for 36%, government owners for 12%, and inventors for 7%. This suggests that while corporate patent holders dominate the field overall, public research and institutional activity play an especially important role in China and South Korea.
When analyzing global patents held by companies and research institutes across the three humanoid robotics technology fields, 10 organizations stand out for the strength of their portfolios, as measured by the Patent Asset Index: Fanuc, Alphabet, Ubtech Robotics, Kawasaki Heavy Industries, Chinese Academy of Sciences, Nvidia, Strong Force Innovation, Harbin Institute of Technology, Intuitive Surgical, and Samsung. They span a diverse set of players, from automation and robotics companies to AI, electronics, surgical robotics, and public research institutions.
The presence of the Chinese Academy of Sciences and the Harbin Institute of Technology is another indicator of China’s strategic focus on robotics research and development.
For nine of the 10 organizations, patents related to the three humanoid robotics technology fields account for less than 10% of their total patent portfolio. This indicates that while they hold highly relevant patents, these technologies represent only a limited share of their broader innovation activity. Kawasaki Heavy Industries, for example, is strongly positioned in industrial robotics and production line automation. Strong Force Innovation appears through patents related to robot fleet management, task allocation, workflow simulation, and the adaptation of robotic resources across value chain environments. Nvidia stands out for patent strength in control and planning architectures, as well as interaction systems, reflecting the importance of AI, computing, and robotics software for autonomous robot behavior. Intuitive Surgical also appears with impactful patents in interaction systems, especially in areas related to precise robotic manipulation for minimally invasive surgery.
Ubtech Robotics is the clear exception. Patents covered by the three classifiers account for more than 25% of its total patent portfolio, showing a close alignment between its technology portfolio and commercial focus on humanoid robots. It is also the only organization in this group that is truly dedicated to humanoid robotics and has publicly demonstrated a bipedal humanoid robot.
The broader patent landscape shows that many of the strongest portfolios are held by organizations active in adjacent robotics, automation, AI, electronics, and research fields. These organizations may own highly relevant patents, but humanoid robotics is often only one part of a much wider technology and business portfolio.
To focus on startup companies where humanoid robotics is central to the business, we applied additional criteria. We analyzed the strength of relevant patent assets held by dedicated humanoid robotic that have publicly demonstrated at least one bipedal walking robot.
The list of most innovative robotics startups was dominated by Chinese-based players, which accounted for six of the ten companies on the list, including Fourier (#1), Agibot (#2), LimX Dynamics (#3), Pudu Robotics (#4), Unitree Robotics (#5), and Leju Robotics (#9). Other top innovators on the list included Agile Robotics (#7), based in Munich, Germany, and three US-based startups, Figure AI (#6) from San Jose, California; Apptronik (#8) from Austin, Texas; and Agility Robotics (#10) from Salem, Oregon.
Ranked by Strength of Related Patent Assets
Source: LexisNexis® PatentSight+™
Founded in 2015, Fourier is a proactive AI robotics company dedicated to enriching human life through full-stack robotic technologies. Fourier has evolved from exoskeleton robots into a diverse portfolio of humanoid robots and intelligent rehabilitation systems. Today, our product lineup — including the GRx humanoid series, RehabHubs, the Galileo system, and a range of open-source solutions — serves more than 2,000 institutions across 40 countries, accelerating the real-world adoption of embodied AI.
Founded in 2023, Agibot is an Embodied AI foundation model company developing both the intelligence layer and the corresponding robotic embodiments needed to bring general intelligence into the physical world. Agibot’s “Three Intelligences in One” architecture integrates Locomotion Intelligence, Interaction Intelligence, and Manipulation Intelligence into a unified embodied system. Its portfolio spans humanoid robots, quadrupeds, dexterous systems, and commercial cleaning solutions.
Founded in 2022, LimX Dynamics is an AI-driven embodied intelligence robotics company driving the innovation of full-size general-purpose humanoid robots and other innovative products. We are committed to disruptive technology in Embodied AI, with the mission to unlock the generalization of Artificial General Intelligence (AGI) in the real world. The company aims to accelerate embodied AI research, development, and real-world deployment across different industries.
Founded in 2016, Pudu Robotics is a global leader in the commercial service robotics sector. Pudu is dedicated to empowering easier work and better lives through AI and robotics, with a vision of building a global intelligent robotics infrastructure that serves 10 billion people worldwide. Pudu Robotics has achieved full-stack proprietary R&D in core technologies, including navigation algorithms, multi-robot scheduling, swarm control, motion controllers, and integrated joint modules.
Founded in 2016, Unitree Robotics is a world-renowned civilian robotics company, which is focusing on the R&D, production, and sales of consumer and industry-class high-performance general-purpose legged and humanoid robots, six-axis manipulators, and other products. Unitree focus on independent research and development and technological innovation, fully self-researching key core robot components such as motors, reducers, controllers, LIDAR and high-performance perception and motion control algorithms.
Founded in 2022 and based in San Jose, California, Figure is an AI robotics company developing autonomous general-purpose humanoid robots. The goal of the company is to ship humanoid robots with human level intelligence. Its robots are engineered to perform a variety of tasks in the home and commercial market
Founded in Munich in 2018 by renowned robotics researchers from the German Aerospace Center (DLR), Agile Robots is a leading provider of next-generation automation solutions. By combining artificial intelligence and robotics, the company makes industries smarter, more flexible, and more efficient.
Apptronik is a human-centered robotics company developing AI-powered robots to support humanity in every facet of life. Our humanoid robot, Apollo, is designed to collaborate thoughtfully with humans—initially in critical industries such as manufacturing and logistics, with future applications in healthcare, the home, and beyond. Apollo is the culmination of nearly a decade of development, drawing on Apptronik’s extensive work on 15 previous robots, including NASA’s Valkyrie robot. Apptronik started out of the Human Centered Robotics Lab at the University of Texas at Austin.
Founded in 2016, Leju is a high-tech enterprise focused on the research of core technologies for robots, as well as the research, development, and production of intelligent robotics products. The company has received strategic investments from Tencent, Shenzhen Capital Group, Hongtai Fund, and others.
Founded in 2015, Agility Robotics’ mission is to build robot partners that augment the human workforce, ultimately enabling humans to be more human. Agility’s groundbreaking general-purpose humanoid robot, Digit, is the first multi-purpose, human-centric robot that is made for work
Other humanoid robotics startups that have received significant media attention finished lower in the ranking based on the strength of their relevant patent assets, including Agility Robotics at #12, DEEP Robotics at #14, and NEURA Robotics at #20. Several other prominent humanoid robotics companies were not included because they do not fit the startup-focused scope of the list, including publicly held UBTECH and Tesla, as well as Boston Dynamics, which became fully owned by Hyundai in June 2026.
UBTECH offers a blueprint for how a humanoid robotics company can grow from a startup into a publicly listed, commercially scaling robotics player. Founded in 2012, UBTECH became the first humanoid robotics company listed on the Hong Kong Stock Exchange in 2023. Its growth reflects a combination of long-term investment in full-stack humanoid robotics technologies, external investor backing, access to public-market capital, and the conversion of industrial humanoid robot pilots into larger commercial deployments. In our analysis, UBTECH holds 370 relevant patent families captured by the three humanoid robotics classifiers, reflecting a close connection between its patent activity and commercial focus. As stated in UBTECH’s 2025 Annual Report, revenue from full-size embodied intelligent humanoid robot products and services increased from RMB 35.6 million, approximately USD 5.1 million, in 2024 to RMB 820.6 million, approximately USD 119.2 million, in 2025, driven by the acceleration of large-scale scenario-based applications and large-scale order deliveries.
The market continues to move quickly, and the next stage of competition may be shaped not only by technology development, but also by financing, public listings, and acquisitions. Some humanoid robotics startups are already moving toward public markets, while others are attracting interest from established technology players. Mentee Robotics is one example. In January 2026, Mobileye announced a definitive agreement to acquire the AI-first humanoid robotics company, illustrating how humanoid robotics startups can become attractive acquisition targets for companies looking to expand into physical AI.
The below chart compares the relevant patent assets of selected humanoid robotics players by portfolio size, shown on the x-axis, average portfolio quality, shown through Competitive Impact on the y-axis, and overall portfolio strength, represented by bubble size and measured by the Patent Asset Index.
This type of visualization can also help model market development and assess how intellectual asset positions may shift through mergers and acquisitions, including how combined portfolios could change a company’s position relative to other market players.
Humanoid robotics is a useful reminder that emerging technology fields rarely fit neatly into one patent class, keyword search, or company list. The most relevant patent activity is often distributed across multiple technical domains. In complex or very niche technology fields, custom classifiers can help capture the field more precisely, especially when standard patent classifications are too broad, too fragmented, or not designed around the application being analyzed.
For IP professionals, this is the bigger takeaway. In application-driven fields, patent analysis should not stop at counting patents or identifying the largest portfolio owners. It should ask a more strategic set of questions: Which technologies define the field? Can those technologies be captured through standard classifications, or is a custom classifier needed? Which patent owners are active across those technologies? How strong are their portfolios? How concentrated is their activity? And does the patent position align with the company’s commercial direction?
This type of layered analysis can help IP teams move from landscape reporting to decision support. It can reveal competitors that are not yet obvious from the market narrative, identify adjacent players with relevant technology positions, distinguish broad patent holders from dedicated specialists, and support more informed decisions around R&D focus, freedom-to-operate monitoring, partnerships, licensing, and portfolio development.
For emerging technologies like humanoid robotics, the most valuable insight often comes from connecting different signals. Patent data provides the technical evidence. Custom classifiers help define the technology field with greater precision. Portfolio strength shows where meaningful innovation assets are being built. Business context shows which companies may be positioned to turn those assets into market impact. Together, these signals give IP professionals a more practical way to understand where a field is heading and where their organization may need to act.
As new technology fields emerge, IP teams should revisit how they define and monitor them. Build patent landscapes around the technical capabilities that make the field possible, not only around familiar keywords or product labels. Use custom classifiers where the field is too complex, niche, or application-driven for standard classifications alone. Combine portfolio size with quality, concentration, and business relevance using patent analytics tools such as LexisNexis® PatentSight+™. Use those insights continuously, not just for one-off reports, but as part of everyday IP strategy, competitive intelligence, partnership evaluation, and innovation planning.
