methodology
How Robot Maids grades household capability
We separate capability, evidence strength and autonomy. A polished manufacturer video can show that a task is possible without establishing that it is reliable.
Capability
RELIABLE means repeated performance in realistic conditions. WORKING means substantial capability with remaining limits. DEMONSTRATED records a meaningful complete or near-complete task. EARLY covers partial or constrained performance. We use MAKER’S CLAIM when the public evidence does not yet justify a stronger label.
Evidence
Independent repeated ordinary-home use is strongest. Repeated real-home deployment follows. Autonomous manufacturer demonstrations are useful evidence; controlled or edited demonstrations are weaker for reliability. We record teleoperation separately.
Robot Maid Score
The current experimental weighting gives 35% to household capability breadth, 25% to evidence/reliability, 20% to autonomy, 10% to domestic practicality and 10% to availability. We show the components beside the total so the number never has to carry the argument by itself.
Housework Automation Index
We will publish the first percentage after we fix the UK household time-use weights and task-maturity rules. We will keep each historical snapshot unchanged.