Stuck in "hand"! Reports indicate that Tesla continues to lower the production target for the "Optimus" robot, which has dropped to 2,000 units this summer

As the core electric vehicle business faces sluggish growth, the humanoid robot project "Optimus," which Elon Musk has high hopes for and claims will eventually surpass the electric vehicle business, is encountering significant setbacks due to key technological bottlenecks.

Media reports, citing two informed sources, indicate that Tesla has abandoned its plan to produce thousands of robots this year due to difficulties in advancing the robot's "hand," the most technically challenging component.

Production targets repeatedly scaled back, "hand" becomes the biggest technical bottleneck

The production ambitions of the Optimus project have experienced a rollercoaster decline since 2025.

At the beginning of the year, Tesla planned to increase production from dozens of units in 2024 to thousands. In March, Musk even raised the target to "at least 5,000 units" during an internal meeting.

However, this aggressive target faced internal resistance from the start. According to informed sources, project employees had repeatedly told Musk that the expansion timeline he set was overly optimistic. Subsequent facts confirmed these concerns: just a few months later, the production target was significantly reduced to 2,000 units.

By summer, employees reported to Musk that due to functional defects in the "hand," even if 2,000 robots could be produced, their practicality would be diminished.

This ultimately led Tesla to abandon its mass production plans for this year and instead redirect resources to tackle the hand design and make other improvements. Musk himself acknowledged in a podcast in September that "the hand, including the forearm, is the main part of the difficulty in the entire robotics engineering."

Accompanying the project delays are changes in the core leadership. Milan Kovac, the head of the Optimus project, left in June.

Currently, the project is led by three executives: Konstantinos Laskaris is responsible for hardware design, Ashok Elluswamy for software, and Lars Moravy for manufacturing, all of whom report directly to Musk. Among them, the latter two executives also oversee Tesla's automotive business.

From "factory worker" to "robot scientist"

Despite the short-term mass production setbacks, Musk's vision for Optimus remains grand. He not only hopes that Optimus can board SpaceX's Mars spacecraft by the end of next year but also plans to deeply integrate it with the artificial intelligence company xAI, ultimately creating a "robot scientist" capable of replacing white-collar workers and even scientists.

Regarding the "Mars plan" for Optimus, media reports cite industry insiders commenting that the current version of Optimus is designed for indoor environments on Earth. To operate on Mars, where the average temperature can drop to minus 80 degrees Fahrenheit (about minus 62 degrees Celsius), it would require a specially designed spacesuit and a complete redesign of its thermal management system to perform any tasks.

As for making Optimus a "robot scientist" conducting research in the lab through xAI's Grok model, this idea remains in the distant conceptual stage of internal discussions Currently, the application of Optimus at Tesla's factory is limited to basic repetitive tasks such as sorting batteries and safety patrols.

Tesla's core logic behind the humanoid design is to leverage billions of human activity videos available on the internet to train the robots. However, transforming these unstructured video data into effective training instructions is itself a significant technical challenge. At this stage, Tesla primarily uses videos of its employees engaged in household chores, walking, and other activities for training.

While Tesla is struggling to overcome technical hurdles, external competition is becoming increasingly fierce. Giants like Meta and Amazon have entered the fray, and startups such as Figure AI and 1X are also securing billions of dollars in funding.

At a summit held in New York last week, when asked about the competitive advantages against giants like Tesla, 1X's CEO Bernt Børnich's answer was succinct:

"The simple answer is a usable product."