What it is
Time-Based Maintenance (the classic preventive): scheduled replacement or restoration by age or usage (hours, cycles, kilometres), regardless of the item's apparent condition.
When it is the right decision (Framework B)
Two simultaneous conditions (Moubray, ch. 6):
- A characteristic failure age — failure probability grows with time/usage: β > 1 in the Weibull analysis, with dispersion low enough for an interval to make sense;
- Restoration returns the original resistance — replacing/overhauling the item genuinely "resets the clock" (true for physical wear; false for operation-induced failures).
The interval comes from Weibull analysis of the history (or the manufacturer's recommendation while data is lacking — revised after 2–3 cycles).
The pitfall that defines the strategy
TBM on a random failure (β ≈ 1) is technically unjustifiable — and worse than useless: every intervention reintroduces infant mortality risk (assembly error, contamination, infant parts) into an item that was not heading toward failure (Nowlan & Heap, p. 46 — the discovery that founded RCM: only ~11% of complex failure modes have a dominant wear-out zone). Before scheduling any periodic replacement, the question is always: does this failure mode have an age? — if not, the path is CBM (if there is a P-F) or RTF.
On the platform
The strategy selector runs this triage question by question; the Weibull calculator estimates β and η from your history — β is the deciding vote between TBM and the alternatives.