Tesla Moved Optimus Training to Dedicated Teams After Workers Objected, Report Says

Tesla has moved movement-recording work for Optimus to dedicated teams after factory employees objected to helping train robots they feared could replace them. The change puts a workforce problem alongside the manufacturing challenges Tesla faces as it tries to increase robot production.

The workers behind the training data

Tesla had factory workers in Texas and California wear suits that recorded their movements as they worked. Those recordings were intended to help teach Optimus manual tasks by giving it examples of human actions, according to The Information’s reporting, described by Ars Technica.

Some employees objected because they believed the robots would eventually replace them. Tesla then assigned the recording work to dedicated teams and established training hubs. That response changes who gathers the examples; it does not remove Tesla’s need for examples of physical work.

The distinction matters: workers’ fear of replacement is reported, but the account does not establish that Optimus has taken their jobs. Tesla is asking people to supply training data for a machine whose ability to perform a broad range of tasks is still limited.

Production has its own bottlenecks

Tesla has reportedly reached production of hundreds of Optimus robots a week and is aiming to exceed 1,000 a week by the end of 2026. The target is not a completed production milestone. It depends on a manufacturing process that is still producing robots in need of immediate fixes.

Tesla has already shifted people toward the robot program. Its Fremont factory stopped making the Model S and Model X in May, and line workers and engineers were reassigned to Optimus work, according to The Information. That reassignment is separate from the objections to wearing recording suits, but it shows how much factory work the program now involves.

The hand assembly problem

The robot’s hands and forearms contain more than 100 small components and still require manual assembly. Some newly built robots need repairs right away. Production equipment has also struggled to line up components precisely, while the line cannot simply run faster without hitting limits, according to The Information’s account.

Touch sensing presents a different problem. After sensors in Optimus hands proved unreliable, Tesla developed a glove-like sensor layer that can be replaced without replacing the entire hand. That design addresses repairability, though it does not by itself solve the difficulty of assembling the hands or raising output.

Training for work the robots cannot yet do broadly

A source cited by The Information described current Optimus robots as needing programming for particular tasks in carefully controlled settings. That is narrower than a robot that can take on varied jobs without being prepared for each one. Human demonstrations may help Tesla train the system, but the recordings are not evidence that Optimus can already perform those jobs on its own.

Tesla therefore faces two different tests on its path to higher output: making the machines consistently enough to meet its production goal, and teaching them useful work. Moving data collection away from factory employees addresses their immediate objection to recording their movements. It leaves the question behind that objection—what role those employees will have if the robots become more capable—unanswered.