MTBF

MTBF (Mean Time Between Failures) is the classic reliability KPI: the average operating interval between two consecutive unplanned failures of a repairable asset. ISO 22400-2:2014 defines it formally as KPI #29 — MTBF = Operating Time ÷ Number of Failures — where 'operating time' is the asset's Actual Production Time (APT) per the ISO 22400-1 time model, and 'failures' are the unplanned stops that consume APT. Units are time per failure (hours/failure is the convention).

The ISO 22400-2 framing matters because plant-floor MTBF numbers are routinely wrong for the same five reasons. (1) Counting calendar time instead of APT — a line that runs one shift a day cannot have its 24h clock in the denominator. (2) Counting planned downtime (changeovers, PM, breaks) as 'time between failures' — only unplanned events count. (3) Mixing failure modes — a sensor swap and a gearbox seizure are not the same failure population, so a blended MTBF hides the real driver. (4) Confusing MTBF with MTTF — MTBF is for repairable assets, MTTF (Mean Time To Failure, KPI #30) is for non-repairable components. (5) Confusing MTBF with MTTR (Mean Time To Repair, KPI #31) — MTBF measures uptime between failures, MTTR measures downtime per failure; both feed Availability = MTBF ÷ (MTBF + MTTR).

MTBF underpins risk-based maintenance, spares stocking, OEE Availability decomposition, and the JIPM/TPM Excellence threshold (85% OEE typically requires Availability ≥95%, which constrains MTBF/MTTR ratio). For GMP plants, MTBF on critical equipment is also an ICH Q9 / 21 CFR 211.67 risk input — assets with deteriorating MTBF feed change-control and PM-frequency reviews.

In V5: every unplanned stop captured at the kiosk becomes a failure event with cause-code, asset ID and APT timestamp; the maintenance dashboard renders MTBF, MTTR and Availability per the ISO 22400-2 formulas, bound to their KPI IDs in the Validation Pack. Trend lines flag deteriorating MTBF before it shows up in OEE.