Report: Tesla's Optimus robot production capacity has increased tenfold, with a target of producing 1,000 units per week by the end of the year.

The mass production of Tesla's humanoid robot Optimus is far more complex than expected. Despite significant increases in production in recent months, technical challenges on the production line, supply chain constraints, and insufficient AI capabilities are hindering the commercialization of this robot, which Musk considers the company's "most important product."

According to sources, Tesla's Optimus production reached several hundred units per week last month, roughly ten times the dozens produced per week during the small-batch testing phase in the second quarter of this year. Management has communicated the goal to employees: to establish a continuous automated production line by the end of the year, achieving a production capacity of over a thousand units per week. However, this goal is still far from the company's final plan of approximately 20,000 units per week. Meanwhile, the Optimus V3 currently rolling off the production line at the Fremont, California factory is not the final version planned for sale to customers; the commercial version must meet more stringent durability and reliability standards.

These challenges directly impact investors' assessment of Tesla's progress in transforming into an AI company. Musk has repeatedly adjusted the Optimus's production timeline, promising a new version to be unveiled by mid-year in April, but that date has passed without a release . He admitted in the July earnings call that the Optimus will be Tesla's most difficult product to scale ever.

Production line issues are concentrated on manual assembly and automated equipment.

One of the core challenges facing the mass production of Optimus lies in the manufacturing process of the robot's hand.

According to sources, the Optimus's hand and forearm contain over 100 screws and other small parts, and currently still require manual assembly by workers. The reliability of the hand's tactile sensors is also questionable. To address this, Tesla has developed a "sensor glove" incorporating tactile sensors that can be replaced individually without replacing the entire hand, and plans to include this system in the production version next year.

At the production line level, multiple workstations responsible for assembling hands and joints, installing software, and testing electronic components all experienced issues with fixtures failing to accurately position parts each time. Optimus components are far smaller than automotive parts, requiring even higher assembly precision, leading to an increase in the number of robots needing rework after production. Some newly developed tooling and automation equipment also failed to operate stably at high speeds on the production line.

In early May of this year, Tesla shut down the Model S and Model X production lines at its Fremont factory, reassigning some workers and engineers to the Optimus test line and drawing dozens of engineers from the Model Y production line to assist in its progress.

The supply chain is highly dependent on China, raising concerns about quality consistency.

At the supply chain level, Tesla faces challenges that should not be underestimated.

Musk pointed out in an earnings call that Optimus's supply chain needed to be built almost from scratch, as the commercial humanoid robot industry has not yet formed a mature supplier network. Currently, while Tesla is developing its own core technologies, it relies on external suppliers for key components such as motors and precision gears that drive joint movements, and the production of these components is largely concentrated in China.

According to sources familiar with the matter, Tesla is testing and certifying new suppliers while pressuring some to build factories outside of China, offering larger orders as an incentive. However, while some suppliers can guarantee quality during the prototype or small-batch stages, they struggle to maintain consistency as production increases.

AI capabilities are insufficient; robots still rely on task-specific training.

Besides hardware manufacturing, Optimus' AI system has also not yet reached the level required for commercialization.

According to three sources familiar with the matter, Optimus currently relies heavily on task-specific training and programming, making its performance unpredictable when faced with untrained tasks or scenarios. Tesla is attempting to build a basic motion library for its robots, combining these basic movements to handle different tasks, rather than training them from scratch for each job. However, it will still take several days for the robots to learn basic tasks.

In terms of data accumulation, Tesla currently has over 500,000 hours of training data and plans to double that number by the end of the year. To this end, the company began establishing training centers in Colorado, Arizona, and Florida this summer, where dozens of data collectors wear helmets equipped with cameras and backpacks carrying computers to collect first-person visual data.

Tesla’s long-term goal is to build a feedback loop similar to its driver assistance software—robots deployed in real-world environments continuously generate data to improve the AI system, and then push the updates to the entire robot corps.

Commercialization path: Leasing first, targeting Tesla-like factories as customers.

In terms of commercialization strategy, Tesla plans to initially provide Optimus to external commercial customers through leasing rather than sales, so that the robots can be recycled, upgraded or refurbished after the lease expires, while continuously optimizing the AI system using data collected from customer factories and warehouses.

Tesla has compiled a shortlist of potential customers, typically targeting businesses with factory or warehouse layouts similar to Tesla's facilities, to reduce the difficulty of adapting the robots to new environments. Musk previously stated that the target price range for the Optimus when it is eventually sold to the public is between $20,000 and $30,000.

Currently, most off-line robots are used for internal testing, training, and data collection, including in tightly controlled areas within Tesla's Palo Alto offices and factory. In these scenarios, the robots are programmed to perform specific tasks, rather than operating as general-purpose or remotely controlled devices. Tesla states that it will gradually expand the robots' operational scope as their performance continues to improve.

Risk Warning and DisclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.