Mean Time Between Failures

平均无故障时间MTBFAdvanced

The average time a repairable machine runs normally between one failure and the next.

Mean time between failures (MTBF) is a standard reliability-engineering metric: the average time a repairable piece of equipment runs normally between successive failures, usually estimated as total operating time divided by number of failures, in hours. It's commonly considered alongside mean time to repair (MTTR); together the two determine what fraction of the time equipment is actually usable. For humanoid and other embodied robots, parts like joint modules, reducers, and dexterous-hand tendons are prone to problems, and a low MTBF means frequent downtime for repairs — one of the metrics factory customers care about most when purchasing and deploying robots. Robot-learning papers more commonly report mean intervention interval (how often a human has to take over), a similar idea but tracking a different statistic.

ExampleIf a robot runs 2,000 hours total and breaks down 4 times, its MTBF is roughly 500 hours.

Also called: MTBF
Related: Mean Time Between Interventions · Number of Nines (Reliability) · Mass Production · Yield Rate / Manufacturing Consistency · Return on Investment / Payback Period
Sources: Mean time between failures - Wikipedia