A standards gap chose the form factor
Humanoid's chief product officer Sotirios Stasinopoulos gave the clearest explanation of the company's design anyone in this field has offered lately: more than 85 or 90 percent of use cases can be covered by a wheel-based platform. That is the commercial half of the argument. The regulatory half is sharper. Humanoid is targeting CE certification for its wheeled HMND 01 next year, and bipedal robots do not have an applicable ISO safety standard to be certified against at all. Wheels are not a compromise on ambition here. They are the only path to a machine a European factory manager can lawfully sign off.
Why it matters: every buyer watching walking-robot demos has been asking the wrong question. The bottleneck on deploying a humanoid into a European plant in 2027 is not whether it can balance, carry or learn. It is whether there is a standard to certify it against, and for legged machines there is not one yet. A company that designed to the standard that exists just became Europe's first pure-play humanoid robotics unicorn, which is the market pricing that fact rather than the technology.
The proof is one shift in Erlangen, not a stage demo
The wheeled model completed an eight-hour autonomous shift at a Siemens factory in Erlangen, moving 60 totes an hour with a pick-and-place success rate above 90 percent. Those three numbers together are worth more than any promotional video, because a full shift is where thermal limits, battery swaps, edge cases and human traffic actually bite, and 60 moves an hour is a rate a plant manager can compare against a person or a conveyor. Nine proof-of-concept projects are complete across logistics, manufacturing and retail, with a tenth running.
Yes, but: a success rate above 90 percent is also the number to interrogate. In a materials handling loop, one failure in ten either needs a human standing by or needs the cell designed so a dropped tote is recoverable without stopping the line. That is a process design cost and it belongs in the business case, not in the vendor's demo. The beta robot arrives in the fourth quarter of 2026 and commercial pilots start alongside it, which means the first honest reliability numbers land about a year from now.
German industry funded its own supply chain
Prime Movers Lab led the round, but the names that matter operationally are Bosch and Schaeffler. Robert Bosch Robotics will act as contract manufacturer, supplying hardware design, production capacity and supply chain expertise, with capacity reported at 100,000 units over five years. Schaeffler has ordered 1,000 robots. Both are also shareholders. Aglaé Ventures, the Arnault family vehicle behind LVMH, and Taiwan's Fubon Financial Holding Venture Capital round out the round, and the founder Artem Sokolov put in roughly 30 million dollars of his own money before any of it.
That structure is how Europe finances hardware when its venture market will not. A pure software company raising 152 million dollars buys engineers. A robot company raising it needs a factory, a supply chain and a first customer, and Europe's answer has been to have the industrial incumbents supply all three and take equity for the trouble. It is a genuine advantage over a comparable American startup that has to buy its manufacturing on the open market. It also means the order book and the production line are the same relationship viewed twice.
What to discount and what to take seriously
Take the certification timeline, the Erlangen shift and the Bosch manufacturing arrangement seriously. Those are checkable, dated and involve counterparties with reputations to lose. Discount the pre-order book. Humanoid claims 34,000 pre-orders worth around 2.5 billion dollars, a figure roughly nine times everything the company has ever raised and one it has not evidenced. Pre-orders on a robot whose beta ships in the fourth quarter of this year are expressions of interest with a deposit at best, and the company has not published conversion terms.
The bottom line: the honest read of this round is a well-run two-year-old company with a defensible form-factor decision, real industrial backing, one credible full-shift trial and an order book that should not be treated as revenue. The team is unusually strong for the age of the business, with more than 50 hires from Boston Dynamics, Sanctuary AI, Apptronik and 1X among roughly 200 engineers. That is a company worth watching and not yet a company worth planning your 2027 headcount around.
Two questions for your own automation plan
First, ask every robotics vendor in your pipeline the certification question directly: which standard, which notified body, what date. Humanoid's answer is a wheeled platform aiming for CE marking next year, and any vendor who cannot match that specificity is selling you a timeline they do not control. This applies with particular force to legged machines, where the answer today is that the standard does not exist, so the deployment date cannot be committed no matter how good the demonstration looks.
Second, check whether your automation supplier is now part-owned by, or manufactured by, a company you already buy from. Bosch and Schaeffler sit on both sides of this deal, and that pattern will repeat as European industrials fund the automation they intend to use. It is not a problem in itself, but it changes your negotiating position, your concentration risk and what happens to your service contract if the relationship between those two companies changes. Ask who owns whom before you sign, not after.
Read next: A Robot Maker Sold 34,000 Units Before Shipping | Kimi's Maker Is About To Publish Its Real Numbers
