The Humanoid Robot Market Is Smaller Than It Looks
None of this argues that humanoid robots will not be important. It argues that the path to importance is narrower and more incremental than the broad market narratives suggest.
Key takeaways
- The most recent headlines on humanoid robots show a surge of activity across both hardware and software fronts.
- Figure AI released a video in May of its Helix 02 humanoids making a bed together, a demonstration meant to prove the machines can handle complex domestic chores such as opening doors, moving furniture and draping clothing. The company, which has raised more than $1 billion and is valued at $39 billion, says the robots have been retrained on new data, but it has not announced a consumer launch date.
- In Shanghai, Chinese startup DroidUp unveiled Moya, a 1.65‑metre, 32‑kilogram “fully biomimetic” robot that mimics human tissue movement with pneumatic muscles rather than traditional motors. Moya can reproduce micro‑expressions, maintain eye contact and walk with 92 percent posture accuracy, and is being positioned for institutions in healthcare and education with a projected commercial release in late 2026.
- Genesis AI introduced its GENE‑26.5 foundation model, a “robotic brain” that captures human hand and finger movements via a glove‑based data system and uses a custom simulation to accelerate training. The model already enables a robot to solve a Rubik’s Cube and perform a wide range of dexterous manipulations, and the company says a first general‑purpose robot based on the technology will be shown soon. Investors such as Khosla Ventures, Eric Schmidt and Daniela Rus have highlighted the breakthrough as a paradigm shift for real‑world AI.
- Hyundai’s robotics programme remains hamstrung by a limited supply of Boston Dynamics’ Atlas units, with the maker reportedly producing only four humanoids per month—insufficient to meet Hyundai’s goal of tens of thousands of robots for its U.S. factories. Boston Dynamics is planning a new manufacturing plant that could move Atlas into a true production version, potentially easing Hyundai’s bottleneck.
None of this argues that humanoid robots will not be important. It argues that the path to importance is narrower and more incremental than the broad market narratives suggest. The early value will come from constrained niches where the physics, economics, and safety can be aligned. From there, capabilities may expand and markets may grow. But the sequence matters. Starting from a $20 trillion labor market and working backward to technology is a recipe for overestimation. Starting from a specific task, with defined objects, environments, and safety requirements, and building outward is how real markets form. The more human-like the task, the more it bundles dexterity, perception, judgment, and safety into a single requirement, and the less attainable it is in the near term. That is the central constraint on humanoid robot markets today. The answer to Zach’s question about who will buy Tesla’s Optimus is likely Tesla itself for its factories, but it won’t need a million of them.
Sign up for CleanTechnica's Weekly Substack for Zach and Scott's in-depth analyses and high level summaries, sign up for our daily newsletter, and follow us on Google News!
Have a tip for CleanTechnica? Want to advertise? Want to suggest a guest for our CleanTech Talk podcast? Contact us here. China’s manufacturing ecosystem can likely compress hardware costs once demand is proven. Tesla’s vertical integration and software stack may offer advantages in system design and deployment. But both face the same underlying constraints. A robot that can move boxes in a warehouse does not automatically become a robot that can do laundry or provide eldercare. Each step up the dexterity and safety axes requires new capabilities, new validation, and new regulatory acceptance.
Mentioned in this article