Sanctuary prioritizes manipulation over walking

This design choice shows that Sanctuary is optimizing for useful factory work and fast learning loops, not for humanoid theater. Phoenix Gen 8 moved to wheels because customers found current bipedal legs too fragile for the heavier, more rigid upper body needed for precise industrial tasks. That fits Sanctuary’s broader hands first strategy, where manipulation quality and data capture matter more than humanlike walking in the near term.

• Sanctuary had earlier argued that legs were important for generality, but the Gen 8 shift reflects a narrower commercial reality. Most early industrial jobs happen at a station, bench, or line, where mobility matters less than stability, uptime, and repeatable arm and hand performance.
  1 sanctuary 2 sanctuary 5 sacra
• The same pattern shows up across peers. Apptronik built Apollo so its upper body can sit on legs, wheels, or a fixed pedestal, because wheels move faster and pedestals give more stability for precision work. The market is converging on task specific lower bodies with a common manipulation stack on top.
  4 sacra 6 sacra
• For Sanctuary specifically, Phoenix is functioning as a research and teleoperation system that generates training data for Carbon and future Physical AI products. A tougher wheeled base means more hours collecting demonstrations, fewer breakdowns, and lower build complexity, which matters more right now than proving elegant walking.
  1 sanctuary 3 sacra 7 sanctuary

The next step is a split product architecture. Wheels and fixed bases will win first in factories because they are cheaper, sturdier, and easier to deploy, while full legs return later for sites with stairs, narrow access paths, or multi station movement. The control stack and hands are becoming the durable core, and locomotion is turning into a swappable module.