Comparable evidence
HumanoidUptime only compares results when metric definitions, tasks and observation conditions align. In the current dataset, no cross-model pair clears that gate.
Comparison gate
A matching headline number is insufficient. Every comparison needs a shared denominator and preserved operating context.
cross-model pairs currently qualify for a performance ranking
Blocked comparisons
Manufacturer-reported cumulative output
Calendar evidence span; runtime undisclosed
Why blockedA cumulative task count and a calendar deployment span have different units, tasks and denominators—even for the same robot model.
Ten-month automotive production period
Warehouse tote-transfer workflow
Why blockedDifferent tasks, output units and missing attempt or assisted-cycle denominators.
Production record; schedule and downtime missing
Single-unit controlled laboratory scenario
Why blockedCumulative operating time and per-charge endurance answer different questions.
Manufacturer product claim
Independent test in a defined scenario
Why blockedA swap-duration claim cannot be compared with measured power draw or endurance.
The useful result
As comparable measurements arrive, this page will show them without changing the admission rules.
Read the methodology