Humanoid robots are leaving the demonstration stage and entering commercial workplaces, as BMW, Mercedes-Benz and Amazon shift from pilot tests to early deployments of general-purpose machines built by startups including Figure AI, Apptronik and Agility Robotics. The transition, accelerating through 2024 and 2025, marks the first time humanoid systems have performed paid, practical work in real industrial facilities. Analysts attribute the shift to converging forces: chronic labor shortages, rapid advances in machine learning and a wave of venture capital now measured in the billions of dollars.
Why Now: The Technology Finally Caught Up
Industrial robots have operated in factories for decades, but they are caged, expensive and programmed for a single repetitive task. Humanoids promise something different: machines that can navigate spaces built for people and switch between jobs with minimal reprogramming.
Three breakthroughs made that possible. Vision-language-action models now let robots interpret their surroundings and respond to natural-language instructions. Training in simulation and from human teleoperation has slashed the time needed to teach new skills. And hardware costs have fallen as electric actuators, batteries and sensors matured.
Nvidia signaled the scale of the opportunity at its GTC conference in March 2024, launching Project GR00T, a foundation model for humanoid robots, alongside dedicated computing hardware. CEO Jensen Huang has described physical AI as the next major wave of the technology.
From Pilot Tests to Production Lines
Figure AI’s Figure 02 robot has been tested at BMW’s Spartanburg, South Carolina, plant, where it handles sheet-metal parts in assembly fixtures. BMW has said the robots are intended for tasks that are ergonomically difficult, tiring or unsafe for human workers.
Amazon has tested Agility Robotics’ Digit in its fulfillment centers since late 2023, moving empty tote bins between stations. Apptronik signed a partnership with Mercedes-Benz in 2024 to pilot its Apollo robot at plants in Germany and Hungary.
Tesla, meanwhile, says its Optimus robot is already performing tasks inside its own facilities. CEO Elon Musk has predicted the humanoid could eventually account for a majority of the company’s value, though production targets have slipped more than once.
The Money Behind the Machines
Investors have moved fast. Figure AI raised $675 million in early 2024 at a $2.6 billion valuation, with backing from Microsoft, OpenAI, Nvidia and Jeff Bezos. Apptronik has raised more than $350 million, including a round with participation from Google.
Forecasts vary widely but point in the same direction. Goldman Sachs Research projects the humanoid robot market could reach $38 billion by 2035, with roughly 1.4 million units shipped. Morgan Stanley has estimated a global market of about $5 trillion by 2050, with more than one billion robots in use.
Skeptics Urge Caution
Not everyone is convinced the timeline is realistic. MIT robotics veteran Rodney Brooks has argued that humanoids will not match human dexterity, reliability or cost-effectiveness for many years, warning that current projections repeat past hype cycles.
The technical hurdles are concrete. Battery life on current models typically runs four to six hours, and uptime in early deployments remains well below human shift patterns. Safety standards for robots working directly alongside people are still being written; no dedicated workplace-safety standard for humanoids yet exists in the United States.
What It Means for Workers and Industry
Manufacturers frame the machines as an answer to labor scarcity rather than a replacement for people. A widely cited Deloitte and National Association of Manufacturers study projects the U.S. could face 2.1 million unfilled manufacturing jobs by 2030.
Labor advocates remain watchful. The AFL-CIO’s Technology Institute has called for workers to have a voice in how AI and robotics are deployed, and unions are pressing for safeguards as machines move closer to human workspaces.
For companies, the strategic calculus is shifting from whether to adopt humanoids to when. Early adopters gain operational data that improves their robots’ models with every shift — an advantage analysts say compounds quickly.
What to Watch Next
The next 12 to 18 months will test whether humanoids can sustain commercial uptime outside controlled pilots. Key markers include Figure’s promised second commercial customer, Tesla’s production targets for Optimus and the first large-volume orders from logistics operators.
Regulation will also take shape. Standards bodies and workplace-safety agencies are expected to issue guidance on human-robot collaboration as deployments widen across North America, Europe and Asia.
If current projections hold, humanoid robots could follow the trajectory of electric vehicles — years of gradual progress followed by rapid scale. The factories of 2030 may feature a few extra sets of hands on the line, and the companies that learn to work alongside them first will hold a measurable head start.