QM BenchQuality engineering tools

FMEA action priority (AP): the AIAG-VDA table explained

How the AIAG-VDA action priority replaces the RPN: what High, Medium and Low mean, why severity comes first, worked S / O / D examples where AP and RPN disagree, and how to use AP in a PFMEA.

Ľubomír Blažek
Process quality team leader, automotive · updated 2026-11-01

Try it while you read

Open the free FMEA Action Priority Calculator tool and click Load sample: it reproduces the case study below.

Open the tool

Theory: why the RPN was replaced

RPN = S × O × D (1 to 1,000)

The risk priority number multiplies three ratings, so very different risks get the same number. 9 × 2 × 2 and 3 × 4 × 3 are both 36, although the first can hurt someone and the second is a minor nuisance. Fixed RPN thresholds also invite small rating changes to get under the limit.

The AIAG-VDA FMEA handbook (2019) uses an action priority table instead: every combination of S, O and D has a priority High, Medium or Low. Severity is considered first, then occurrence, then detection, so a high severity with a realistic occurrence cannot be averaged away.

PriorityMeaning
HighAction required to improve prevention or detection, or a justification why the current controls are adequate
MediumAction should be taken, or the reason why not documented
LowAction could be taken

Key point

Lowering occurrence (prevention) is almost always better than improving detection: a defect that is never made does not need to be found.

Step by step

  1. Rate S from the effect on the customer and the end user, O from the prevention controls and history, D from the detection controls, with the AIAG-VDA tables.
  2. Look up the AP in the action priority calculator; it also shows the RPN for comparison and the full table for that severity.
  3. Decide the action for High and Medium, with owner and due date.
  4. Re-rate after the action, usually O and D only; severity changes only with a design change.
  5. Work the full FMEA in the FMEA tool, which applies the same table to every row.

Case study: where AP and RPN disagree

FailureSODRPNAP
Missing weld nut, found by camera92236Low
Wrong torque, checked by audit only947252High
Bore oversize, SPC every hour756210Medium
Cosmetic scratch, visual check588320High
Label smudged34336Low
Brake line clip not seated1035150Medium

With a fixed RPN limit of 200, the team would work on the scratch (320) and the torque (252), and ignore the brake line clip (150) because it is below the limit. The action priority keeps the torque High, flags the scratch too because occurrence and detection are both very poor, and puts the safety-relevant clip at Medium, so it gets an action instead of disappearing under the threshold.

Result

For the torque the team adds a torque-angle monitoring with automatic stop (D from 7 to 3) and a better tool (O from 4 to 3): the new AP is Low.

Common mistakes

  • Changing ratings until the AP drops, instead of improving the process.
  • Lowering severity with an action that only improves detection.
  • Rating detection low because “the customer will find it”.
  • High priorities without an action and without a documented justification.

More background: RPN and action priority and DFMEA vs PFMEA.

Frequently asked questions

Is the RPN still allowed?

The AIAG-VDA FMEA handbook (2019) replaces the RPN with action priority. Many automotive customers require AP; outside automotive the RPN is still common. Some teams show both during the change.

Why can a high RPN have a low priority?

Because the RPN treats S, O and D as equally important. A failure with severity 5 but occurrence and detection of 8 has RPN 320, yet the effect is moderate. A severity 9 failure with RPN 36 can still need attention. AP weighs severity first, then occurrence, then detection.

Does AP High mean the process is not allowed?

No. It means the team must take an action to lower the risk or document why the current controls are adequate, and management is aware of it.