PRELIMINARY - IDEAL-FIXTURE SIMULATION ONLY - NOT APPROVED FOR CONNECTION, POWERED TESTING, MOTION, WALKING, OR ENERGIZATION

The whole robot now runs in a real physics engine.

MuJoCo compiles the 9.928 kg tether-first model and executes both 10.72 s sequences with a numerical six-degree-of-freedom trajectory fixture. This is a model-integration test—not proof of free walking.

Bounded results

WS-L01

1.11°

Maximum rotary tracking error; saturation 84.7%; declared-support coverage 100.0%.

FAIL IDEAL-FIXTURE NUMERICAL SCREEN

WS-R01

1.11°

Maximum rotary tracking error; saturation 84.6%; declared-support coverage 100.0%.

FAIL IDEAL-FIXTURE NUMERICAL SCREEN

Scrub recorded physics data

Loading simulation samples…

Download the evidence

sequence summary · all 25 axes · 50 Hz simulation samples · interactive preview data · derived MuJoCo model · runtime provenance · open holds

What the ideal fixture hides

The fixture can apply arbitrary six-axis constraint load to the pelvis path. Consequently this run cannot establish free balance, recoverability, physical tether forces, sole friction, contact compliance, real actuator tracking, current or thermal capacity, stopping behavior, or safety. Its value is narrower: the dynamics tree compiles, mass/inertia is positive, contact topology is explicit, torque-limited numerical control executes, and failures are measurable.