If Elon Musk is known for his grandiose, boundary-pushing ambitions, then the humanoid robot “Optimus” is surely one of his boldest bets. Remember Tesla’s 2022 AI Day? That clunky prototype, which could barely walk without stumbling, didn’t stop Musk from making grand claims on stage that it would change how humans live and be the secret to pushing Tesla’s valuation to $25 trillion.

Many laughed it off as pure hype, but there’s no denying Tesla has actually improved this “metal oaf” quite a bit in the last two years. By the October 2024 “We, Robot” event, Optimus could autonomously navigate to its charger, play rock-paper-scissors with humans, and even serve drinks gracefully—like something straight out of a sci-fi movie.

Musk’s plan is anything but modest: he envisions a world with 20 billion humanoid robots, and wants Tesla to capture at least 10% of that market. According to his roadmap, Tesla will first produce a few thousand units for in-house testing in 2025, ramp up to 50,000–100,000 units in 2026, and aim for over 500,000 units per year by 2027, with unit prices in the $20,000–30,000 range to make them affordable for regular households.

But while the vision is grand, reality is harsh.

Rare Earth Bottleneck: Tesla Hits a Supply-Chain Wall

During an April 22 interview, Musk openly acknowledged that scaling up production faces a significant hurdle: securing enough rare-earth magnets.

China, aiming to curb military uses of rare earths, has tightened export controls. Rare-earth magnets are essential for making high-performance motors and sensors—the critical components of robots. Without them, you can’t build efficient, lightweight, durable motors, which means forget about robots reliably hauling goods in factories or serving drinks at home.

In other words, Optimus is now facing a major global supply-chain choke point.

Team Turmoil: Project Head’s Emotional Resignation

And as if that weren’t enough, more trouble hit.

On June 7, Milan Kovac, who led Tesla’s humanoid robotics program, announced his resignation on X with an emotional note, calling it “the toughest choice of his life” and explaining he needed to return overseas to be with family. He emphasized that his departure was for personal reasons and unrelated to other matters.

But few bought that official line. Just three days earlier, Tesla had announced it was kicking off trial production of Optimus at the Fremont factory, aiming to build 5,000 units in 2025, and bragged that it had already ordered parts for 12,000 robots.

Kovac wasn’t an outsider brought in to take the blame. He’d been with Tesla since 2016, playing a key role in Autopilot development. Under his leadership, the robot program evolved from an early, clumsy prototype to a more advanced second-generation version capable of moving materials, performing poses, and imitating human actions. His departure quickly raised worries about possible disruption in Tesla’s robotics research and development.

Production Delays, Parts Orders on Hold

And those concerns weren’t unfounded.

Insiders say Tesla suddenly hit the brakes on parts procurement for Optimus to do a round of hardware and software redesigns—a pause expected to last about two months. By late May, Tesla had purchased enough materials to build around 1,200 robots but had only assembled fewer than 1,000—falling noticeably short of its planned pace.

Hardware Shortcomings: Overheating Joints, Limited Battery Life, Unfinished Dexterous Hands

Beyond supply-chain issues and team turmoil, Optimus itself still has some very real “body problems.”

Reports say some joint motors are prone to overheating, the dexterous hands can’t yet handle heavy loads well, and the drivetrain components have less-than-ideal lifespans. Battery life is also limited, meaning it can’t roam around the house doing chores all day—it might need to go recharge before it can even finish serving a meal.

Tesla is reportedly testing multiple technical solutions, shopping for dexterous-hand suppliers, and evaluating samples from different vendors. No one doubts Musk will keep throwing money at the problem, but making something both cheap and truly useful in the short term remains a tall order.

The Future of Humanoid Robots: Slow and Steady Wins the Race

The bottom line is that building a robot that can “walk a few steps” isn’t all that hard. But making one that can be mass-produced and reliably work in factories, homes, or care settings is another matter entirely.

Tesla does have strengths in AI, perception, control, motors, and energy management. But solving all the hardware bottlenecks, overcoming supply-chain challenges, and tackling ethical concerns and cost pressures will take time. The much-touted “robot revolution” will almost certainly arrive more slowly than Musk claims.

Optimus is at a crucial crossroads right now. It needs a stable team, sustained R&D investment, and steady progress on hardware, software integration, and cost control. If Tesla can pull it off, it might deliver a genuinely world-changing product within the next decade. If it fumbles, Optimus could go down as just another overhyped “moonshot” that fizzles out.

Either way, Musk won’t give up easily—and we can all keep watching to see how big a vision he’ll sell next, and how much of it he can actually deliver.