Tesla has pulled the plug on two of its best known cars to make room for a robot. The Fremont factory in California stopped building the Model S and Model X in May 2026, and both line workers and engineers there were moved onto the Optimus humanoid robot program instead.
Part of that new job involves teaching a robot how to move like a person. Tesla has had factory workers wear motion capture suits and cameras to record their movements so Optimus can learn from them, a method the company also uses for its self driving cars. Some employees pushed back once they realized what the data was for, telling reporters they knew the robots were built to eventually replace them. Tesla responded by moving that recording work to separate, dedicated teams and setting up specific training hubs, rather than asking regular line workers to do it.
The bigger problem may be the robot itself. Each Optimus hand and forearm contains more than one hundred small parts that still have to be assembled by hand, and the sensors inside them were unreliable until Tesla built a replaceable, glove like layer to fix it. Weekly output has climbed from a small test run to several hundred units, and Tesla wants to pass one thousand robots a week by the end of the year. Even at that pace, most robots stay inside Tesla's own factories, running narrow, preprogrammed tasks in supervised areas rather than doing general work on their own.
This is not the first time an Optimus timeline has slipped. Musk said in 2022 that production could start the following year, then promised ten thousand robots for 2025, a number that never materialized. By January of this year he admitted that none of the robots built so far were doing anything genuinely useful. Betting markets currently give Optimus a low chance of reaching an actual consumer release by the end of this year.
The part of this story that matters most for other businesses is not the robot, it is the workers. Tesla's approach of recording employees' bodies to build their own automated replacement is a preview of a tension that will show up anywhere companies try to automate physical work: people are being asked to help build the tool that removes their job, and that is a trust problem before it is a technology problem. Any company planning automation should expect this friction and plan for it directly, rather than assuming staff will cooperate quietly.
There is also a bigger competitive story sitting underneath Tesla's struggles. China's humanoid robot makers, led by Unitree and AgiBot, are already shipping thousands of units and are on track for national production to grow roughly five times over in a single year, at a much lower cost than Tesla's hand built approach. Tesla is also still leaning on Chinese suppliers for robot parts, even as it tries to build robots domestically. If cheap, mass produced humanoid robots do eventually reach factories and warehouses at scale, the company setting the price is more likely to be based in China than in Fremont.