Tesla Starts Optimus Robot Production at Fremont as Profit Drops 43% on Heavy AI Spending

Reviewed byNidhi Govil

15 Sources

Share

Tesla launched production of its Optimus humanoid robot at its Fremont, California factory, marking a shift from electric vehicle manufacturing to AI-driven robotics. The company's Q3 2026 earnings revealed a 43% profit decline to $787 million as operating expenses surged 39% to fund AI research, autonomous driving, and robot development despite flat revenue of $28.04 billion.

Tesla Launches Optimus Production Amid Sharp Profit Decline

Tesla has begun manufacturing its Optimus humanoid robot at the Fremont, California factory, converting production lines previously dedicated to electric vehicle manufacturing. The company announced during its Q3 2026 earnings that "the bots coming off this line are the first production-quality bots that we have manufactured," signaling a major shift in its manufacturing strategy

1

. However, this milestone comes as Tesla's Q3 earnings showed a 43% profit drop, with net income falling to $787 million from $1.37 billion in the same quarter of 2025

4

. Revenue remained flat at $28.04 billion, slightly below the $28.09 billion generated in Q3 2025

1

.

Source: TechCrunch

Source: TechCrunch

Operating Expenses Surge as AI Investment Accelerates

Tesla's operating expenses climbed 39% to $4.78 billion, driven primarily by $2.76 billion in research and development spending focused on AI, autonomous driving, and the Optimus robot

4

. Capital expenditures more than doubled to $4.73 billion, with Chief Financial Officer Vaibhav Taneja projecting total capital spending of approximately $20 billion for the year

4

. Operating income dropped 56% to $718 million, representing a margin of just 2.6%

4

. The company has secured borrowing capacity of up to $30 billion, though it does not plan to draw on these funds this year

4

.

Automotive Revenue Stalls While Services Grow

Tesla delivered 486,532 vehicles in Q3, approximately 6,000 more than the previous quarter but lower than the same period in 2025

1

. Automotive revenue slipped 3% to $20.56 billion, remaining flat year-over-year at $20 billion

5

. Energy revenue fell 16% to $2.88 billion, which Tesla attributed partly to tariff-related customer refunds

4

. Services revenue provided a bright spot, growing 33% year-over-year to $4.6 billion, boosted by increased subscriptions to Full Self-Driving advanced driver-assistance software

1

. The services segment generated a record $734 million in gross profit for the quarter

1

.

Source: Japan Times

Source: Japan Times

Major Engineering Challenges Remain for Optimus

During the earnings call, Elon Musk acknowledged substantial work remains before Optimus becomes commercially viable. "There's still quite a lot of work to do with regard to the AI and the robot for it to be truly useful for open-ended tasks," Musk stated

2

. Critical challenges include perfecting the robot's hands, ensuring safety, reliability, and durability, and developing the AI needed for productive work

2

. "We're building robots that do not have perfected hands," Musk admitted, highlighting the difficulty of creating hands that can withstand repeated use while remaining sensitive enough to handle delicate objects

2

. Ken Goldberg, a robotics professor at the University of California, Berkeley, warned that "if you have a lot of hardware without the software, I think that's a big problem"

2

.

Optimus Academy to Generate Training Data Through Self-Play

Tesla plans to build between 10,000 and 20,000 Optimus robots for what Musk calls "Optimus Academy," where robots will practice tasks to generate training data through a process known as self-play

3

. "In order for Optimus to get the training data we need, we actually need to make a lot of Optimi," Musk explained during the earnings call

3

. Unlike Tesla's vehicles, which provide vast amounts of real-world driving data, Optimus has no existing fleet to learn from

2

. Musk compared the process to raising a child: "They initially are sort of wobbly and they don't know how to do things, and eventually they learn to wipe their own ass, which is a big milestone"

3

. Ashok Elluswamy, who leads Tesla's AI software team, noted that robots would initially learn by watching humans perform tasks

3

.

Production Ramp Expected to Be Slow and Bumpy

Musk described the current phase as "early production" and warned that scaling would follow an S-curve with a slow, bumpy beginning

2

. "There's like 10,000 things that have to go right for production of the robot, and it will move as fast as the slowest," he stated

2

. Most equipment for a production line designed to eventually manufacture one million robots annually is already installed in Fremont

2

. However, reaching a rate of 20,000 robots per week will take "at least a year, maybe a little longer," according to Musk

5

. Parts of the robot's design are still changing, requiring Tesla to build an extensive global supply chain and expand domestic factory capacity

2

.

Source: The Next Web

Source: The Next Web

Commercial Delivery Timeline and AI5 Chip Integration

Tesla expects to begin delivering Optimus to external companies around the end of 2027, though the company did not specify which customers or what tasks the robots would perform

2

. Initial applications will focus on dangerous, repetitive industrial work that requires around-the-clock operation

3

. The AI5 chip, expected to provide approximately 10 times the computing power of the current AI4, should be integrated into Optimus around April 2027

3

. Musk said Tesla could release a video within the next month or two demonstrating Optimus performing non-pre-programmed tasks

3

. "I remain convinced that Optimus will be the biggest product on Earth," Musk declared, adding that "at some point there will be more humanoid robots on Earth than there are humans"

3

.

Expert Skepticism on Timeline and Technical Feasibility

Robotics experts remain skeptical about Tesla's aggressive timeline. Goldberg noted that while robots have become adept at walking, running, and jumping, dexterity remains significantly harder. "Manipulating flexible materials, such as wires and fabric, or assembling products is incredibly difficult; doing it quickly and reliably enough for real work is harder still," he explained

2

. "I've been working on these problems for 40 years. I don't think they're going to be solved in the next year or two," Goldberg stated

2

. Some of his colleagues believe a truly useful humanoid robot remains 25 years away, though Goldberg himself expects progress sooner than that—just not on Musk's timeline

2

. Tesla is not alone in pursuing AI-driven robotics for industrial applications; a competing humanoid robot recently completed an eight-hour shift at a Siemens factory in Germany

3

.

© 2026 TheOutpost.AI All rights reserved