Mean Time Between Failures (MTBF) is the average time a repairable asset runs between failures, calculated as total operating time divided by the number of failures. Higher MTBF means greater reliability.
Together with MTTR it determines equipment availability. MTBF is improved through preventive and predictive maintenance and TPM.
Example. A conveyor ran 720 hours in a month and failed 4 times. MTBF = 720 / 4 = 180 hours. The four repairs took 2, 1.5, 3 and 1.5 hours, so MTTR = 8 / 4 = 2 hours. Inherent availability = MTBF / (MTBF + MTTR) = 180 / 182 = 98.9%. Raising MTBF to 360 hours lifts it to 99.4%; halving MTTR to 1 hour gets to 99.4% as well, and is often the cheaper move.
Count failures consistently: a failure is a stop that needs maintenance intervention, not a minor stop the operator clears. Track MTBF per asset and by failure mode, because a machine with a 180-hour MTBF driven by one sensor is a different problem from one with ten unrelated causes. The maintenance guide covers PM intervals set from MTBF data.