The Factory Floor Gets Its Electric Motor Moment
The Factory Floor Gets Its Electric Motor Moment
When electric motors first arrived in factories, owners did the obvious thing: they ripped out the steam engine and bolted the new motor onto the existing machine. The floor layout did not change. The workflow did not change. Productivity gains were modest — until someone had the radical idea of redesigning the factory itself around what electric power actually enabled.
That history is the best lens for evaluating what Humanoid announced this week.
The British robotics startup, incorporated as SKL Robotics Ltd. and founded in 2024 by Artem Sokolov, has signed a binding deal to deploy between 1,000 and 2,000 wheeled humanoid robots at Schaeffler's German manufacturing plants by 2032 Reuters. Phase one begins in December 2026 at Herzogenaurach, where the robots will handle boxes inside a live production environment, and at Schweinfurt, where a three-month capability demonstration and integration test will precede three months of on-site validation aimed at "stable, continuous operation approaching full production scale" Humanoid press release. The agreement is structured as Robot-as-a-Service: Humanoid provides the robots, fleet management software, maintenance, and 24/7 support. Schaeffler pays by the month.
On its own terms, this is a significant disclosure. A two-year-old company with roughly $50 million in founder-led capital has secured a contract with one of Germany's largest industrial suppliers Forbes — not a pilot, not a letter of intent, but a binding phased deployment. Schaeffler CEO Klaus Rosenfeld has separately told Reuters that the company expects its humanoid robotics order book to reach "hundreds of millions of euros" by 2030, based on collaborations with roughly 45 humanoid robotics players globally and five active customer contracts Reuters. Schaeffler is not a fringe buyer. It is a tier-one automotive and industrial supplier with the balance sheet to make this a credible reference installation.
But the history of factory electrification offers a specific, testable prediction about what happens next: the first generation of deployments will underdeliver relative to the hype, because the humans deploying the robots will slot them into unchanged workflows rather than redesign those workflows around what the robots can actually do.
The Schaeffler phase-one scope confirms this pattern. Box handling is not a fundamental rethinking of manufacturing labor. It is the most tractable version of the problem — repetitive, physically constrained, measurable. The Schweinfurt validation period is explicitly about proving the robots can hold up in an operating environment, not about inventing new ways of organizing production. Humanoid's own press release describes the initial goal as "meeting Schaeffler's requirements for system architecture, safety, and IT infrastructure" and "aligning with standardized rollout processes" Humanoid press release. That is a retrofit, not a redesign.
The actuator supply agreement is where the real production ambition hides. Schaeffler will become Humanoid's preferred supplier, covering more than 50% of the company's joint actuator demand for its wheeled platforms through 2031, in a deal expected to total at least one million actuator units Reuters. Forbes did the math: at roughly 20 Schaeffler-supplied actuators per robot, 50% coverage implies something approaching 100,000 robot units over the five-year term Forbes. Humanoid's CEO Artem Sokolov described the Schweinfurt proof-of-concept, which involved the HMND 01 Alpha — a 220-centimeter wheeled industrial humanoid unveiled in September 2025 after a reported seven-month development cycle — as showing "strong results" Forbes. But the gap between a proof-of-concept and 100,000 deployed units across a five-year horizon is not a straight line.
Schaeffler is not putting all its eggs in one basket. The company is simultaneously buying at least 1,000 humanoid robots from Switzerland's Hexagon Robotics and is party to a partnership with Germany's Neura Robotics that could total 4,000 to 6,000 of Neura's 4NE1 models Forbes. This multi-vendor approach is not a sign of confidence in any single robot maker. It is a hedge — a tier-one industrial buyer acknowledging that the humanoid robotics market is still immature enough that picking a single winner would be foolish.
The Robot-as-a-Service framing is the most commercially significant aspect of the deal, and the part most likely to determine whether humanoid robotics scales beyond well-capitalized early adopters. Moving automation from capital expenditure to operational expense removes the upfront cost barrier that has historically limited robotics adoption to companies large enough to amortize the investment over decades. If a mid-size manufacturer can subscribe to a robot fleet the way it subscribes to cloud compute, the addressable market expands dramatically. That is the theory. The counterargument is that RaaS providers carry the deployment risk — if the robots do not perform, the provider eats the cost — which creates pressure to low-ball the technical scope and overpromise on capability. Schaeffler is getting box handling and a validation period, not a transformation.
The factory floor has been waiting for this moment for a long time. The question is whether anyone is actually ready for what it means.