Tesla has already deployed more than 1,000 Optimus units across its Fremont and Texas gigafactories, performing tasks ranging from battery assembly to quality inspection, marking a significant milestone in humanoid robot adoption within manufacturing. The deployment underscores a broader industrial shift: humanoid robots are moving from research labs into real production environments, where they're handling complex, repetitive work alongside human teams.

How Are Humanoid Robots Being Trained for Factory Work?

The path from prototype to production floor involves a structured training methodology that major manufacturers are now standardizing. Hexagon Robotics and Schaeffler have partnered to establish what they call a "Humanoid Gym" in Germany, where AEON robots learn industrial tasks through imitation learning. This approach allows robots to repeat actions until they achieve reliable performance, then validates their capabilities against real production scenarios before full deployment.

• Imitation Learning: Robots learn by observing and repeating human actions, building muscle memory through repetition until they can perform tasks reliably without human intervention.
• Real-World Validation: Training occurs in dedicated facilities that mirror actual factory conditions, allowing engineers to test robot performance against genuine production challenges before scaling deployment.
• Phased Deployment: Companies begin with pilot programs in controlled environments, then gradually expand to full manufacturing lines once robots demonstrate consistent accuracy and reliability.

Schaeffler's partnership with Hexagon targets deployment of at least 1,000 AEON units across its global manufacturing sites in the coming years. The company plays a dual role: it's both a customer testing the robots and a critical component supplier, providing the actuator technology that enables the robots' movement. This arrangement creates a feedback loop where real-world performance data directly informs hardware improvements.

What Tasks Are Humanoid Robots Actually Performing in Factories?

The work humanoid robots are handling today spans a range of manufacturing operations that were previously difficult to automate. Tesla's Optimus units are working across battery assembly, EV pack loading, cable routing, connector seating, parts handling, sorting 4680 battery cells, kitting, quality inspection, and pick-and-place operations. BMW has logged more than 1,250 operational hours with Figure AI's Figure 02 humanoid at its Spartanburg plant, achieving over 99% placement accuracy while loading sheet metal parts into welding machines.

Mercedes-Benz is using Apptronik's Apollo robots at its Berlin Digital Factory Campus to transport components to production lines, perform initial quality checks, and deliver kitted parts. These deployments demonstrate that humanoid robots aren't limited to a single task; they're capable of learning multiple operations and adapting to different production workflows. The consistency of these results across different manufacturers and robot designs suggests the technology has reached a maturity threshold where industrial deployment is now economically viable.

According to Counterpoint Research, 16,000 humanoid robots were installed globally in 2025, a figure that reflects the accelerating pace of adoption. The manufacturing sector faces significant workforce pressures: Deloitte and The Manufacturing Institute project that the U.S. manufacturing sector alone will need to fill 3.8 million jobs by 2033, with up to 1.9 million potentially going unfilled if current trends continue. Humanoid robots are emerging as a partial solution to this labor shortage.

Why Is Rare Earth Supply the Real Bottleneck in the Robot Boom?

While artificial intelligence captures headlines, the physical infrastructure powering humanoid robots depends on a critical material that's concentrated in a single country's hands. Every humanoid robot contains dozens of electric motors packed into its shoulders, wrists, hips, and ankles, and each motor requires rare earth magnets, chiefly neodymium-iron-boron compounds. A single humanoid can carry two to four kilograms of rare earth magnets, sometimes exceeding the amount found in an entire electric vehicle.

China produces roughly 90% of the world's finished rare earth magnets and has already demonstrated its willingness to use this leverage as a geopolitical tool. During this year's tariff standoff, China restricted exports and blacklisted U.S. firms, creating supply chain vulnerabilities that could constrain the global robot boom. The parallel to the 1973 oil embargo is instructive: a critical dependency concentrated in a rival's hands becomes a weapon waiting to be deployed.

Washington has begun responding to this vulnerability. The Pentagon has taken a 15% stake in MP Materials, the one U.S. company capable of mining, refining, and producing finished rare earth magnets at scale. Defense rules now require contractors to phase out Chinese magnets by January 2027. Apple and General Motors have both signed magnet supply deals with domestic producers, signaling that supply chain resilience is becoming a strategic priority across industries.

"Whoever supplies the magnets supplies the empire," noted Luke Lango, editor of Innovation Investor at InvestorPlace.

Luke Lango, Editor, Innovation Investor

The market opportunity is staggering. Morgan Stanley estimates the humanoid robot market could reach $5 trillion by 2050, with a robot workforce exceeding 1 billion units. Citibank projects the market could reach $7 trillion. Elon Musk has called Tesla's Optimus "the biggest product of all time," estimating long-term revenue at $10 trillion with eventual demand exceeding 20 billion units. These projections underscore why control over rare earth supply has become a strategic priority for governments and corporations alike.

The convergence of AI maturity, manufacturing labor shortages, and supply chain vulnerabilities is creating a moment of inflection for humanoid robotics. Tesla's deployment of over 1,000 Optimus units, Schaeffler's commitment to 1,000 AEON robots, and the broader industrial adoption across BMW, Mercedes-Benz, and other manufacturers signal that humanoid robots have transitioned from experimental technology to operational reality. The next phase of growth will depend not on advancing artificial intelligence, but on securing the physical materials that make these machines move.