What it is
Run-to-Failure: the deliberate decision to operate the item until functional failure and then correct it — with a spare on the shelf, a replacement procedure ready and the failure consequence accepted in advance. It is planned corrective maintenance; the opposite of corrective-by-surprise.
When it is the right decision (Framework B)
Two simultaneous conditions:
- Random failure without P-F — β ≈ 1 (no age to justify TBM) and no monitorable signal to anticipate the failure (no window for CBM). The canonical example: an electronics board burned by a surge.
- Tolerable consequence — the failure is evident and its cost (downtime, repair, safety) is acceptable against the cost of any prevention attempt that, by definition, would not work.
If the consequence is NOT tolerable (safety/environment, or economically intolerable), RTF is forbidden and the compulsory way out is redesign — redundancy, better component selection, elimination of the failure mode (Moubray, ch. 9; the last line of the decision diagram).
What RTF demands (it is not "doing nothing")
- Spares management sized by λ and lead time;
- A fast-replacement procedure (MTTR is the variable left to manage);
- Failure recording in the CMMS with a standardized code — the history is what will allow reclassification if a pattern emerges.
On the platform
The strategy selector reaches RTF through the "no P-F + evident + economic consequence" path — and compulsory redesign when the consequence is safety.