The latest Optimus V3 design remains hard to manufacture consistently. Production equipment has struggled to align components precisely, and running the line at higher speeds can introduce defects. Some newly built robots reportedly need repairs almost immediately after assembly.
The hands are a particularly difficult part of the process. Each hand-and-forearm assembly includes more than 100 small components, including screws, and workers still assemble them manually.
Tesla has also run into reliability problems with the robots’ touch sensors. The company responded by developing a replaceable, glove-like sensor layer so failed sensors can be swapped without replacing the entire hand.
Those hardware issues matter because Tesla is trying to move quickly from low-volume production toward a much larger weekly output. Reaching the company’s year-end target would require both faster assembly and fewer defects as production accelerates.
The software side is not yet fully general-purpose either. One source cited by The Information described current Optimus deployments as requiring programming for specific tasks in carefully controlled settings.
Training the robots has created a separate workforce problem. Tesla asked employees at factories in Texas and California to wear motion-capture suits while working so the company could record their movements for imitation-learning data.
Some employees objected because they “knew the robots were designed to eventually replace them,” according to a worker cited by The Information.
Tesla has since moved more of that collection work to dedicated teams and established training hubs staffed specifically for gathering robot-learning data.
The company also remains dependent on Chinese suppliers for a number of Optimus components. That reliance comes as the broader U.S. robotics industry faces pressure to build more of its supply chain domestically.
In July, the Federal Communications Commission banned new foreign-made robots in several categories, including humanoid robots, four-legged robotic systems and robot vacuum cleaners.
Tesla is also entering an increasingly crowded market. Toyota plans to spend billions upgrading factories with more robotics, including humanoid systems, while Hyundai expects to deploy as many as 25,000 Atlas humanoids from Boston Dynamics over the next several years.
Agility Robotics has already moved further into commercial deployment. The Oregon-based company began using its humanoids at a GXO Logistics warehouse near Atlanta in 2024.
Even with those deployments, the economics and safety of humanoid robots at large scale remain unsettled. Companies are still working through questions around cost, reliability and how robots behave around human workers.
For Tesla, those questions are now tied directly to a major manufacturing push. The company has shifted employees and factory capacity toward Optimus, but its current production process still relies heavily on human assembly and rework.
The near-term challenge is straightforward: Tesla needs to make Optimus faster, more reliably and with less manual intervention if it wants to move from hundreds of robots per week to more than 1,000. The same workers helping build and train the machines are also confronting what those machines could eventually mean for their own jobs.
This analysis is based on reporting from ars TECHNICA.
Image courtesy of Elon Musk/X.
This article was generated with AI assistance and reviewed for accuracy and quality.
