1. Humanoid robots are the story everyone’s chasing. The components are the actual trade.
This account is all about that gap. Between a flashy demo and a real shipped robot sit a dozen critical layers - actuators, screws, bearings, magnets, sensors. That’s where the real value and scarcity live. One simple rule: buy the joints, not the robot
@humanoidsdaily If the brain really is portable across embodiments - G1 humanoid and a stationary dual-arm from one model - then the software stops being anyone's moat and the differentiation drops back down to the hardware:
i.e who builds reliable actuators and joints cheapest, at volume
@tphuang Localizing assembly is the (relatively) easy 50%. I wonder if BYD does ever move LFP cathode/precursor out of China, or is that the permanent moat that stays home no matter how high tariffs go?
4. So I don’t waste time asking which robot wins. I ask: Which layer is genuinely scarce? Who makes it? What share do they control? And where has the market mispriced that reality?
1. Humanoid robots are the story everyone’s chasing. The components are the actual trade.
This account is all about that gap. Between a flashy demo and a real shipped robot sit a dozen critical layers - actuators, screws, bearings, magnets, sensors. That’s where the real value and scarcity live. One simple rule: buy the joints, not the robot
3. That means the scarce components hold the real pricing power, not the company doing the final assembly. The hardest part in the entire robot is the planetary roller screw.
Optimus needs around 14 of them, and three European suppliers control roughly two-thirds of the precision market. You don’t replace decades of specialized thread-grinding expertise with a purchase order and a few quarters.