When will the humanoid patent wars break out?
A spate of stories in the trade press and a recent Wall Street Journal op-ed, “Humanoid robots are the next threat from China” by Senator Katie Britt and Jacob Helberg highlight an urgent need for the US to take action against the rising influence of China in humanoid robotic technology.
The emphasis on the critical need for North American leadership in this sector affecting both military and many civilian industries follows 2024 reports from Goldman Sachs that recent advancements in artificial intelligence (AI), machine learning (ML) and other factors will propel the global humanoid market to 70% annual growth through 2035. It could potentially reach $154 billion by the end of 2035 in a blue-sky scenario (that would be approximately one-third of the 2021 global smartphone market).
With the deep tech involved in humanoid robotics R&D and the well-chronicled patent wars that accompanied the rise of the smartphone, we thought it timely and prudent to look at the patent situation in humanoid technology. We note that there have been reports from numerous sources about China’s “lead” in robotics patents, while others indicate a rapidly closing race.
In this article we provide data and insights into the evolving patent landscape of humanoid robots with an emphasis on North American and European granted patents and applications. Both Tech+IP and Own Innovation have been actively working with leading innovators in the humanoid robotics market for some time. We believe that this review is important both from the point of view of what it says about innovation and what it might portend from a competitive standpoint into the future.
If we are to believe Senator Britt, this kind of patent landscape is even more important due to the limited right to exclude that patents represent for the innovative companies that own them (and even in some instances government agencies such as NASA and others).
Two technologies differentiating humanoids
Although we note that publication of patent applications and patent issuances is only a partial proxy for actual innovation and potential market leadership, we think that with proper search techniques and caveats, it can still be a useful proxy, at least for directional or trend information. We anticipate various improvements in future versions of this study.
In performing this preliminary analysis, we focused on issued US, Canadian and European patents and published patent applications as of April 2024. We left out patents issued or pending in China and other jurisdictions for a host of reasons, including that assessing quality of such patents is quite difficult and as shown with smartphones and other technologies, the patent wars may involve Chinese companies, but the main battlegrounds are typically in these jurisdictions rather than inside China. Future, more refined analyses should consider additional jurisdictions (as has also occurred in smartphone and other standard essential patent licensing).
Importantly, we also note that a humanoid robot uses numerous technologies and structures in common with non-humanoid robots that have been the subject of R&D for at least 20 years and that comprise well over 100,000 patents (72,618 robotic patents being granted between 2005 and 2019 alone). Medical robotics is a substantial area of patenting activity that may show up in certain searches related to humanoid robots, we excluded medical robotics from our study.
To efficiently get to the heart of the matter, we decided to focus for the time being on two main groups of technologies that both differentiate humanoids from other kinds of robots and that in substantial part have been propelled by technological advances in applications of AI and ML (including large language model (LLM) technologies and the use of numerous technologies to propel training and adaptation as well as execution).
Technology groups in our study include:
The first group encompasses general-purpose robots (GPRs), particularly their training features and execution capabilities (Figure 1).
The second group encompasses dexterous manipulation technologies (including hardware and software) due to their importance in emulating human elements such as the human hand, thumb and fingers that work together to grasp, hold and manipulate objects of all kinds (Figure 2). The ability of the human thumb to oppose other fingers, for example, is considered a crucial evolutionary adaptation allowing development of human dexterity and complex tool use.
These developments have significantly influenced innovations in the field and thus the patent landscape, contributing to the future landscape of robotics. While we concede there are other areas of substantial importance as well (and welcome community feedback), this narrowing helped us to choose both differentiated technologies for humanoids and isolate humanoid robotics patents from other associated technologies having applications across multiple industries like automotive, manufacturing, logistics and warehousing.
We note, of course, that being restricted to these two main groups necessarily excludes some highly innovative robotics companies that also hold patents, reflecting their investment in innovation. This includes large numbers of patents from companies such as Hyundai (now with Boston Dynamics), Microsoft, Mujin, Samsung and Tesla.
Insights and trends
Using patent classification methods and sematic techniques in a recursive process to run, test, identify and improve search results, the Tech+IP and Own Innovation teams identified over 3,000 assets relating specifically to General Purpose Robot training features and execution capabilities and nearly 1,000 assets relating to dexterous manipulation technologies (including hardware and software). Results were then stack-ranked by the owning company in each area – we ranked the first 15 in each category in Figures 1 and 2 below, respectively.
Figure 1: Patent holders in general-purpose robotics technology
Figure 2: Patent holders in dexterous manipulation technology
Twenty total companies are represented in the top 15 patent-holding companies in the two segments, while 10 companies are represented in each segment’s top 15. North American firms represent 11 of the 20 firms (Canada 1, USA 10) while Japanese firms represent five of the 20. Korea, Germany and China come in at two, one and one of the 20 companies, respectively.
A comparative analysis between the two groups shows that the growing number of patents in the first group highlights a stronger focus on innovations related to execution and training capabilities.
For example, Alphabet has created a diverse IP portfolio encompassing more than 450 US, CA, and EP patents and pending applications, potentially giving it a significant advantage over other companies in the field. This concentration of inventions reflects significant advancements in their dexterous management capabilities.
Hyundai both on its own and owing to its control of Boston Dynamics has a top three position in both GPRs and dexterous manipulation. Vancouver, Canada-based Sanctuary AI – while a new name possibly to some readers of this article, but very well known for its innovation and patenting – holds a position in these two sectors that is one of the most important globally and that positions it extremely well for a $100 billion-plus global market in the 2030s.
While China does appear to lead in the deployment of robotic technology in industry, the data tells a different story in terms of patent assets outside of China. Ubtech Robotics from Shenzhen, China is the 10th-ranked company in GPR patents, yet no other Chinese companies make the top 15 in GPRs. The next highest-ranked Chinese companies is Midea Group Co, ranked 54th in GPR and 24th in dexterous manipulation technologies. Of the top 100 companies, only three are of Chinese origin while others are spread across geographies, led by North America.
As far as recognised startup and emerging growth companies focused on humanoid robotics are concerned, only Canadian company Sanctuary AI and Brain Corp, headquartered in San Diego, make the list. Companies such as Agility Robotics, Figure, Apptronik and 1X (and others) are recognised in industry trade press as highly innovative; however, these firms do not yet appear to have patent portfolios that match their reputation for innovation. Note that that due to Hyundai’s 80% ownership and SoftBank’s 20% stake, we did not calculate Boston Dynamics on a stand-alone basis.
Conclusion
Although there has been significant publicity surrounding the emergence and potential dominance of Chinese companies in the field of humanoid robotics, an examination of patents in this area – both as a proxy for innovation and as a future marker for competitive issues – reveals that Chinese firms do not hold a significant position. As the humanoid robotics market continues to grow and evolve, however, the role of Chinese companies within this landscape will likely shift considerably as well.
Using the smartphone industry as a guide, Chinese companies were initially not a factor outside of China, held few patents, and were often playing defence versus patent-rich companies from the US, Europe, Japan and Korea. Unable to “invent backwards”, these companies turned to acquiring patents from non-Chinese third parties with early R&D; next they invested heavily in R&D and patenting of future generations of the technology. That strategy turned Huawei, for example, from a net payor of patent royalties to a business yielding $500 million-plus in patent licensing royalties annually.
As the saying goes, if you do not heed history you are doomed to repeat it. It is likely that the lessons of smartphones will be repeated in humanoid robotics, and it behooves companies and investors in the robotics industry to take patenting seriously – as a C-suite and board-level issue – and not subject to quarterly revenue pressure.
To do otherwise introduces unnecessary (and potentially existential) business risk right at the time the market will be exploding. For policymakers, rather than focusing industry-by-industry, it is equally important to take prompt action to reverse the erosion of patent rights on a global scale. If this trend is not reversed across all strategic industries, when patent rights are needed to protect businesses built on innovation in the robotics space, nothing will be left except an empty shell.
Patent owners with comments or questions are invited to contact us to dive deeper into the methodology and results.