Framework B — Prescription

With the diagnosis in hand, Framework B decides what to do — in the RCM order of preference (SAE JA1011; Moubray, RCM II). A task only enters the plan if it is technically APPLICABLE (detects or prevents that failure mode) and WORTH DOING (costs less than the consequence it avoids).

The task hierarchy

1. 1
   CBM

   On-condition — if the failure gives a measurable warning (P-F interval), monitor the condition and act on the signal. Inspection ≤ P-F/2.

2. 2
   TBM

   Scheduled restoration/discard — only when a defined wear-out zone exists (β > 1 with predictable age). Interval before the wear zone.

3. 3
   Proof Test

   Failure-finding — for HIDDEN functions (protections, redundancies): tests the function at the interval (FFI) dictated by the required availability.

4. 4
   RTF

   Run to failure — legitimate when nothing above applies and the consequence is economic only.

5. 5
   Redesign

   Redesign — compulsory when the consequence is safety/environmental and no task reduces the risk to a tolerable level.

Consequence as the filter

The same technical failure changes strategy with its consequence: safety/environmental demands proven risk reduction (or redesign); operational weighs downtime cost; non-operational weighs repair cost only. That is why the selector asks about consequence before closing the decision.

In the collection

Every failure-mode note publishes with its Framework B decision and an exportable maintenance plan (green panel) — coherence enforced at build time: CBM requires a P-F interval, proof testing requires a hidden failure, RTF requires no P-F.