FMEA and control plan guide: RPN, special characteristics and reaction plans

Every quality problem you have ever had was, at some earlier point, a line in an FMEA that either was not written or was not acted on. Failure Mode and Effects Analysis is the discipline of listing how a part or process could fail, scoring how much each failure matters, and deciding what to do about the ones that matter most. The control plan is where that decision becomes concrete — the sheet the line actually works to. Together they are the difference between hoping quality happens and engineering it.

This guide is for quality and process engineers working to IATF 16949 or ISO 9001. It explains the FMEA and the control plan as one connected pair, with a worked, illustrative example. For the wider context — how these two sit among all five automotive core tools — see the pillar guide, What is quality management software?

1. What are the FMEA and the control plan?

FMEA — Failure Mode and Effects Analysis — is a structured, team-based method for identifying the ways a product or process could fail, the effects of each failure, its causes, and the controls in place to prevent or detect it. Each failure mode is scored so the team can prioritise action on the ones that carry the most risk.

The control plan is the document that turns the FMEA's conclusions into daily practice. For every characteristic that must be controlled it states what to measure, with which gauge, at what frequency, by what method — and, crucially, the reaction if a reading is out of specification. It is the bridge between the risk analysis and the shop floor.

The reason they belong together is the special characteristic: a feature the FMEA flags as high-risk becomes a controlled line on the control plan, is measured by inspection and watched by SPC, and — if it fails anyway — drives an 8D whose corrective action re-scores the FMEA. See FMEA & Control Plan and Inspection & SPC.

2. DFMEA vs PFMEA

There are two FMEAs on most automotive parts, and confusing them is a common source of gaps. They ask different questions about different failure sources.

Both use the same scoring model and both flow into the same downstream artefacts, but the PFMEA is usually the one that drives the control plan, because the control plan is a manufacturing document — it controls the process, operation by operation. Where the supplier is design-responsible, the DFMEA additionally shapes the drawing and the special-characteristic list. The two should be consistent: a special characteristic identified in the DFMEA must appear in the PFMEA's controls and on the control plan.

3. Severity, occurrence, detection and RPN

The engine of an FMEA is three ratings, each on a 1-to-10 scale, applied to every failure mode / cause row.

• Severity (S) — how serious the effect of the failure is, from 1 (no discernible effect) to 10 (affects safe operation or violates a regulation without warning). Severity is a property of the effect and the design — inspection cannot reduce it.
• Occurrence (O) — how likely the cause is to occur, from 1 (failure eliminated through prevention) to 10 (failure almost inevitable). Occurrence is reduced by design or process changes that attack the cause.
• Detection (D) — how unlikely the current controls are to catch the failure before it escapes, from 1 (control will almost certainly detect) to 10 (no control, or the control cannot detect). Better detection — mistake-proofing, in-process gauging — lowers D.

The three multiply into the Risk Priority Number: RPN = S × O × D, a value from 1 to 1000 used to rank which failure modes need action first.

Two cautions that separate a real FMEA from a scoring exercise. First, severity is never reduced by adding inspection — only a design or process change lowers it — so a high-severity failure (say S = 9) deserves action even at a modest RPN, and simply detecting it better is not enough. This is why the newer AIAG-VDA method replaces raw RPN ranking with an Action Priority (High / Medium / Low) table that weights severity first; the S, O and D inputs are the same, but the prioritisation no longer lets a low-occurrence, high-severity item hide behind a small RPN. Second, an RPN is only useful relative to the others in the same FMEA — there is no universal "safe" threshold.

4. Special characteristics

A special characteristic (also called significant or critical, with customer-specific symbols) is a product or process feature whose variation materially affects safety, regulatory compliance, fit, function or customer satisfaction. It is the most important output of the FMEA, because it decides what the control plan must control most tightly.

A special characteristic earns extra treatment throughout the system:

• Identified in the FMEA — typically where severity is high — and marked with the customer's symbol.
• Carried onto the control plan with a tighter tolerance, a higher inspection frequency, and often a mistake-proofing or SPC control method rather than a spot check.
• Prioritised in SPC — its capability (Cp, Cpk) is monitored continuously, not just verified once at PPAP.
• Traceable end to end — an IATF audit will follow one special characteristic from the FMEA, to the control plan, to the inspection and SPC results, to confirm the chain is unbroken.

If that flow breaks anywhere — a characteristic flagged in the FMEA that never reaches the control plan, or a control plan feature never actually measured — the whole exercise loses its value. Keeping the link intact is far easier when the FMEA, the specification master and the control plan live in one system rather than three spreadsheets.

5. Building the control plan

A control plan is a table with one row per characteristic, and its columns are its discipline. Miss a column and you have a checklist, not a control plan.

There are three control plans across a part's life, matching the APQP phases: a prototype control plan, a pre-launch control plan (tighter, higher frequency, before the process is proven), and the production control plan. Each characteristic's gauge should come from the gauge register so it is one that passed MSA and is in calibration; each specification should come from the specification master so inspection and SPC read the same limits. The reaction plan column is the one most often left thin — and it is the one that matters most at 2 a.m. when a reading is out and the operator needs to know exactly what to contain and whom to call.

6. A worked example (illustrative)

To make it concrete, here is an illustrative PFMEA row and the control-plan line it produces for a single characteristic on a machined shaft. The numbers below are illustrative, for teaching only.

Because severity is 8, the team cannot inspect the risk away — they act on occurrence and detection. The improvement: add automatic tool-wear offset (drops occurrence) and move to a bore/plug check every 5th part with an SPC chart on the diameter (drops detection). That produces the following control-plan line, and a lower RPN.

With detection improved to 3 and occurrence to 2, the illustrative RPN falls to 8 × 2 × 3 = 48 — and, just as important, the line now has a written reaction that contains the problem to a five-part window instead of an hour's production. This is the shape of a real deployment such as Kakade Laser or Shree Engineering.

7. Keeping it a living document

The most common failure of FMEAs is that they are written once for PPAP and never touched again. A control plan and FMEA earn their keep only if they change when reality does.

8. How Fast Quality Software runs FMEA and control plans

Fast Quality Software is the automotive QMS of the Fast Suite, built in Pune by Improsys under the Fast Technology brand and deployed cloud or on-premise. It keeps the FMEA and control plan connected to the specification, the gauge and the corrective-action loop:

9. Frequently asked questions