Deployment
The strongest evidence is not the demo, but pilot conversion, uptime, cost per task, and repeat customers.
Why it matters
The strongest evidence in humanoid robotics is not a viral video. It is deployment at a customer site.
The question is how many robots are doing which tasks, for how many hours, with how much human intervention, and whether the customer wants more.
Core metrics
- Paid pilot vs. unpaid showcase
- Pilot conversion into fleet orders
- Robots deployed per site
- Hours worked and uptime
- Cost per task
- Human intervention rate
- Repeat orders or expanded scope
Bottleneck hypothesis
Many companies can build impressive robots. Far fewer can deploy them repeatedly, integrate them into customer workflows, prove ROI, and maintain them in the field.
Deployment and integration may become the most important filter as the category matures.
Reference
Humanity’s Last Machine — Ch. Component Costs, Skeleton, Batteries, Compute, PCBs. The BOM cost chapter breaks down which subsystems are structurally expensive vs compressible at scale. Useful for calibrating deployment cost expectations and understanding margin pressure points as production ramps.
Watchlist seed
Figure x BMW, Apptronik x Mercedes, Agility in warehouse/logistics, Boston Dynamics x Hyundai, UBTECH industrial deployments, and Symbotic warehouse automation are evidence tracks to maintain.