The first robot this company plans to sell in volume has no legs

Here's the deal: on July 21, London-based robotics company Humanoid announced a $152 million Series A at a $1.35 billion post-money valuation, bringing total capital raised to $270 million. The company describes itself as Europe's first pure-play humanoid robotics unicorn. It was founded in 2024, so it got there in two years.

Prime Movers Lab led the round, joined by Schaeffler, Bosch, Fubon Financial Holding Venture Capital and Aglaé Ventures. Schaeffler and Bosch are the names to focus on. Both are large German industrial component manufacturers, which makes them potential customers and partners rather than purely financial investors.

They are already behaving that way. Humanoid has an agreement with Schaeffler for large-scale deployment of thousands of humanoid robots in manufacturing environments, which the company calls the largest publicly announced commercial agreement in the industry. Schaeffler CEO Klaus Rosenfeld said the company is "actively backing one of the most exciting frontiers in technology."

The most interesting detail sits elsewhere in the roadmap, though. The first model heading for mass manufacturing is Alpha Wheeled — the wheel-based robot. There's a bipedal model too, Alpha Bipedal, but wheels go into the factory first. That single choice compresses the company's entire strategy.

The cast: Artem Sokolov and 250 engineers

Humanoid was founded by Artem Sokolov in London in 2024. In this announcement he said the company had gone "from an idea to becoming Europe's first pure-play humanoid robotics unicorn" in just two years. It now employs more than 250 engineers, researchers and innovators, with offices in London, Boston, Vancouver and San Diego. It's introduced as a European company, but the footprint deliberately sits inside North America's robotics talent clusters as well.

The technical core is KinetIQ, the company's proprietary AI framework — described as a four-layer "AI brain" that lets robots understand situations, reason about them, and execute complex physical tasks. This is where the real competition in humanoids is happening now. Hardware capability has converged across several companies; what differentiates is how well a robot copes with situations nobody scripted.

Schaeffler plays more than an investor role in this story. It manufactures bearings and motion components for automotive and industrial markets and operates large production facilities worldwide. For a humanoid robot company, that's close to an ideal first customer: high volumes of repetitive logistics and assembly work, capital and experience in automation, and above all a controlled environment where a robot can fail safely.

Bosch's participation follows the same logic. Add partnerships with SAP, Nvidia and Siemens and the picture resolves. Humanoid is not aiming at the consumer market or general-purpose household robots — it chose a route straight into European manufacturing supply chains. That's a deliberate contrast with US competitors like Tesla's Optimus and Figure.

The numbers and the schedule

What was announced, laid out:

Item Detail
Round Series A, $152M
Post-money valuation $1.35B
Total raised $270M
Lead investor Prime Movers Lab
Strategic investors Schaeffler, Bosch, Fubon Financial Holding VC, Aglaé Ventures
Founded 2024 (London)
Headcount 250+ engineers and researchers
Offices London, Boston, Vancouver, San Diego
Products Alpha Wheeled, Alpha Bipedal
AI stack KinetIQ (four-layer framework)
Beta deployment From Q4 2026 in logistics, manufacturing, retail
Mass manufacturing Wheel-based model first

Strategically the last two rows matter most: beta robots going into real sites from Q4 2026, and volume manufacturing starting with the wheeled variant.

Bipedal locomotion is the symbol of humanoid robotics, and in industrial settings it's mostly unnecessary. Factory floors are flat. Warehouse floors are flat. Legs matter for stairs and irregular terrain — environments where you're unlikely to deploy robots anyway. Meanwhile legs substantially increase cost, power draw, failure rates and safety risk. Wheels extend battery life, increase payload, and eliminate the possibility of falling over.

So shipping wheels first under the name "humanoid" isn't a compromise, it's a practical judgment. The human-shaped upper body — two arms, hands, cameras at human height — is what lets the robot use tools and workstations designed for people. Locomotion is simply a place where wheels win. Several competitors are arriving at the same conclusion.

What each party gets

Humanoid gets capital and proving grounds. Robot hardware is expensive to develop, and standing up a production line requires capital of a different magnitude again. The $152 million is runway to the next stage, and the Schaeffler agreement means there's a landing zone at the end of it. Most robotics startups die exactly in the gap between "the demo works" and "someone bought it," and this company is attempting to bridge that gap with a contract.

Schaeffler and Bosch get priority access. European manufacturing labor costs keep rising and skilled-worker shortages are structural. If humanoids actually work, whoever deploys first gets a cost advantage. Taking equity is how you secure that priority — and it also buys a seat at the table when robot specifications get written.

For investors, a deep-tech fund like Prime Movers Lab putting $152 million into a Series A says something about capital temperature in robotics. A $1.35 billion valuation on a two-year-old company is narrative-based rather than revenue-based. What separates it from pure narrative is that concrete demand-side evidence — the Schaeffler agreement — is attached.

For Europe, there's industrial policy significance. Humanoid robotics competition has been led by the US and China. Europe dominated the previous generation of industrial robot arms through KUKA and ABB, but has been faint in this one. European industrial incumbents investing directly in a homegrown startup reads as an attempt to close that gap.

Precedents: what worked and what didn't

Robotics startup history offers plenty of cautions. The canonical failure is Rethink Robotics, founded by Rodney Brooks, which built the Baxter and Sawyer collaborative robots and was genuinely pioneering. It ended in an asset sale in 2018. The cause was economics, not technology — the robots were slower than people, and what factories actually wanted was throughput rather than flexibility.

The counter-example is Universal Robots, the Danish company that skipped flashy humanoids and built a market around easily programmable collaborative arms. It worked because the task scope was well-defined and payback periods were short. When a customer recovers the cost within six months, the purchase decision gets easy.

The humanoid generation has no success story yet. Agility Robotics' Digit has run trials in Amazon warehouses and Figure has piloted at BMW plants. Both are meaningful progress and neither is volume deployment. Tesla's Optimus has generated far more announcement than externally verifiable deployment.

Together those cases point at the criteria that will decide Humanoid's fate. First, payback period — how many years of labor cost the total cost of ownership of one robot recovers determines real adoption. Second, uptime; working in a demo and running three shifts for six months are entirely different problems. Third, the actual terms of the Schaeffler agreement, since "thousands of robots" means something very different as a firm order than as staged options.

How the competition responds

American competitors are ahead on capital. Figure has raised substantially larger rounds repeatedly, and Tesla brings its own manufacturing capability plus battery and motor supply chains. Agility Robotics has accumulated the most real-world logistics deployment experience. Against those, Humanoid's $270 million is not large. What it has instead is proximity to European manufacturing customers, and whether that offsets the capital gap is the open question.

China's counterplay is price. Unitree and its peers already ship humanoids at far lower price points, and supply chain advantages make that gap hard to close. But for European and American manufacturing customers, data security and supply chain risk weigh heavily in purchase decisions, so price alone won't settle the market. Putting Schaeffler and Bosch on the cap table is Humanoid's defense on exactly that axis.

The reaction of incumbent industrial robot makers is another variable. ABB, KUKA and Fanuc already own factory automation, with overwhelming customer relationships and service networks. If they pursue humanoid form factors seriously, the space for startups narrows. If they move slowly to protect existing product revenue, that delay is the opening companies like Humanoid need.

Nvidia sits where it always sits — selling infrastructure to every participant. By providing robot foundation models and simulation platforms, it profits regardless of who wins. Nvidia appearing on Humanoid's partner list reflects that structure more than any special relationship.

So what actually changes

If you work in manufacturing or logistics, Q4 2026 is the date to watch. When beta robots enter real sites, we get the first operational data rather than controlled demos. Uptime, mean time between failures, and how human-robot collaboration is actually organized will give you a basis for judging when this technology becomes usable. Planning deployments now is premature; building an internal list of which tasks would suit a humanoid is not.

For robotics developers, the software stack is the more interesting thread. If a four-layer framework like KinetIQ gets published or partially written up, it becomes a reference point for how to separate robot foundation models from execution layers. There's no standard here yet, every company is answering differently, and the field learns each time one of those answers becomes public.

For investors, three things to track. The specific terms of the Schaeffler agreement — firm order quantities and delivery schedules would let you test the valuation. The production cost of the wheeled model, since unit price relative to labor cost determines addressable market size. And the timing and terms of the next round, because hardware companies burn capital fast and Series B pricing usually reflects market assessment more honestly.

At an industry level, this round signals that humanoid robotics has moved from "when will this work" to "where will it work first." The current answer is: factories. Not consumer household robots but controlled industrial environments with repetitive tasks and a computable payback period. Shipping wheels first is the same answer stated differently.

🥄 Three Things You're Probably Wondering

— Does a two-year-old company deserve $1.35 billion? Not on revenue. What distinguishes it from pure hype is attached demand-side evidence, particularly the thousands-of-units agreement with Schaeffler. Real validation comes from the Q4 2026 beta deployments.

— Why wheels if it's a humanoid? Factory and warehouse floors are flat, so legs add nothing there while substantially raising cost, power consumption, failure rates and safety risk. The human-shaped upper body is what lets the robot use existing workstations and tools, so you keep that and swap the locomotion. It's the rational compromise right now.

— Can it compete with Chinese manufacturers? On price, hard. Chinese firms build far cheaper on supply chain advantage. But European and American manufacturing customers weigh data security and supply chain risk alongside price, and Humanoid built its defense on that axis by taking Schaeffler and Bosch as shareholders. Too early to call the winner.

Sources

Numbers and criteria are as of announcement and may change. Investment calls are yours to make!