FMEA RPN and Action Priority calculator
Rate severity, occurrence and detection from 1 to 10. You get the risk priority number (RPN) and the AIAG-VDA Action Priority (High, Medium or Low), with the reason. Free, no sign-up, and nothing you type leaves your browser.
Action Priority per the AIAG & VDA FMEA Handbook (2019) table for DFMEA and PFMEA · Updated 15 Nov 2026
FMEA worksheet
A short worksheet that ranks rows by Action Priority, then severity, then RPN. It starts with the worked example below; edit any cell, add your own rows, and download a CSV for Excel. Rows are kept in this browser only.
| Failure mode | Effect | Cause | S | O | D | RPN | AP | Remove |
|---|---|---|---|---|---|---|---|---|
| Bearing not fully seated (short press stroke) | Noise and early bearing failure in the field | Press stop worn or set wrong | 7 | 5 | 7 | 245 | High | |
| Bearing pressed in tilted (cocked) | Bearing races damaged, rough rotation | Worn locating pin on fixture | 7 | 3 | 5 | 105 | Medium | |
| Housing bore oversize (loose fit) | Outer race spins in the housing, wear | Boring tool wear not compensated | 8 | 4 | 3 | 96 | Medium | |
| Scratch on housing face | Cosmetic mark some customers see | Parts stacked without separators | 3 | 7 | 8 | 168 | Low |
Worked example: why RPN and AP disagree
The worksheet above is a process FMEA (PFMEA) for one assembly step: pressing a ball bearing into an aluminium housing. Look at two rows.
- Scratch on housing face: S 3, O 7, D 8 gives an RPN of 168, the second highest on the sheet. Its Action Priority is Low. A cosmetic mark that happens often and is hard to spot is a nuisance, not a reason to divert the team.
- Housing bore oversize: S 8, O 4, D 3 gives an RPN of only 96, the lowest on the sheet. Its Action Priority is Medium. A spinning outer race kills the product, so the table keeps it on the action list even though the in-line bore gauge usually catches it.
With an "act on RPN above 100" rule the team would work on the scratch and skip the bore. AP reverses that.
Now the top row, the bearing that is not fully seated (S 7, O 5, D 7, RPN 245, High). Suppose the team adds force-and-displacement monitoring to the press, which rejects any part that does not reach the seated position. Detection drops from 7 to 2: RPN 70, AP Medium. Adding a hard mechanical stop that cannot wear out of position also cuts occurrence to 3: RPN 42, AP Low. Severity stays 7 throughout: only a design change (for example a shoulder in the housing that the bearing seats against) can lower it.
The AIAG-VDA Action Priority table
Each cell is the Action Priority for a severity band (rows), an occurrence band and a detection band. Severity 1 is always Low. Of the 1,000 possible S/O/D combinations, 318 are High, 214 Medium and 468 Low.
| Severity | Occurrence | Detection | |||
|---|---|---|---|---|---|
| 7–10 | 5–6 | 2–4 | 1 | ||
| 9–10 | 8–10 | H | H | H | H |
| 6–7 | H | H | H | H | |
| 4–5 | H | H | H | M | |
| 2–3 | H | M | L | L | |
| 1 | L | L | L | L | |
| 7–8 | 8–10 | H | H | H | H |
| 6–7 | H | H | H | M | |
| 4–5 | H | M | M | M | |
| 2–3 | M | M | L | L | |
| 1 | L | L | L | L | |
| 4–6 | 8–10 | H | H | M | M |
| 6–7 | M | M | M | L | |
| 4–5 | M | L | L | L | |
| 2–3 | L | L | L | L | |
| 1 | L | L | L | L | |
| 2–3 | 8–10 | M | M | L | L |
| 6–7 | L | L | L | L | |
| 4–5 | L | L | L | L | |
| 2–3 | L | L | L | L | |
| 1 | L | L | L | L | |
| 1 | 1–10 | L for every detection rating | |||
H = High, M = Medium, L = Low. This is the table for design and process FMEAs; FMEA for monitoring and system response (FMEA-MSR) uses a different table and is not covered here. Transcribed from public reproductions of the 2019 handbook table and checked cell by cell; if your customer has its own table, theirs wins.
RPN vs Action Priority: what changed in 2019
For decades the default FMEA ranking was the RPN: multiply the three ratings and work down the list from the top. The 2019 AIAG & VDA FMEA handbook, which merged the American (AIAG) and German (VDA) automotive methods, replaced it with the Action Priority table. Four problems with RPN, each visible in the numbers:
- Equal weights. S = 10, O = 2, D = 5; S = 2, O = 5, D = 10; and S = 4, O = 5, D = 5 all give RPN 100. The first could hurt someone; the second is a minor annoyance. AP rates them Medium, Low and Low.
- Thresholds mislead. On the AP table, 21 combinations with RPN below 100 are High (the lowest is RPN 54: S 9, O 6, D 1), and 55 combinations with RPN of 100 or more are Low (up to RPN 210: S 3, O 7, D 10).
- Gaps in the scale. RPN runs from 1 to 1,000 but only 120 values can occur, so the numbers look more precise than they are. An RPN of 240 is not "twice as risky" as 120.
- Detection gets gamed. The cheapest way to cut an RPN is to add inspection and lower D. AP gives detection the least weight, so it pushes teams towards prevention (lower O) and design changes (lower S) instead.
Action Priority is not a risk score. It ranks the need for action, and you cannot sort or add AP values the way people did with RPN. If you need an order inside the High group, sort by severity, then occurrence, which is what the worksheet above does.
How to rate severity, occurrence and detection
A short summary of what each band means for typical design and process FMEAs. These are our paraphrases, grouped the way the AP table groups them. Your customer's or your company's rating tables, or the full tables in the AIAG & VDA handbook, take precedence.
Severity (S): how bad is the effect?
| Rating | Level | What it looks like |
|---|---|---|
| 9–10 | Safety or regulation | The effect could injure someone (user, operator) or break a law or regulation. 10 = no warning, 9 = with warning. |
| 7–8 | Primary function lost or badly degraded | Product does not do its main job. In a plant: a large share of output scrapped, or the line stops. |
| 4–6 | Secondary function or comfort affected | Product works but something customers notice and dislike. In a plant: rework off-line or a slower station. |
| 2–3 | Minor annoyance | Appearance, noise or feel noticed by some customers. In a plant: small rework in station. |
| 1 | No noticeable effect | Nobody can tell the failure happened. |
Occurrence (O): how likely is the cause, given the prevention controls?
| Rating | Level | What it looks like |
|---|---|---|
| 8–10 | Very high | New design, technology or process with no history, or no prevention control at all. 10 = practically certain to happen. |
| 6–7 | High | Similar to designs or processes that had problems; prevention controls only partly effective. |
| 4–5 | Moderate | Mostly proven design or process with some changes; prevention controls usually work. |
| 2–3 | Low | Proven design or process; prevention controls (standards, simulation, mistake-proofing of the cause) are known to work. |
| 1 | Extremely low | The cause is eliminated by design: it cannot physically occur. |
Detection (D): how likely are the controls to catch it before the part leaves?
| Rating | Level | What it looks like |
|---|---|---|
| 7–10 | Low to almost none | No test, a test method that is not proven, or random visual checks. 10 = the failure cannot be detected before it reaches the customer. |
| 5–6 | Moderate | A proven test or inspection, but manual (visual, gauge by an operator) or not run on every part. |
| 2–4 | High | Proven, automated detection on every part (sensor, in-line gauging, end-of-line test) that stops bad parts. |
| 1 | Almost certain | The failure mode cannot be made, or is always caught by design or error-proofing. |
Three rating rules that save arguments in the review meeting:
- Rate severity for the worst effect of the failure mode, and keep it the same everywhere that effect appears. Inspection never lowers severity.
- Rate occurrence for the cause with the prevention controls in place today, not the ones planned.
- Rate detection for controls that exist and are proven. "Operator will look for it" on every part is still a manual check (moderate at best).
Skill-Lync's design programs teach DFMEA and PFMEA on industry projects alongside CAD, GD&T and design for manufacturing.
- Post Graduate Program in CAD (covers DFMEA and PFMEA)
- PG Program in Complete Passenger Car Design & Product Development
- Medical Product Design & Development (FMEA for product risk)
- Design for Manufacturability (DFM), 5 weeks
Frequently asked questions
How do you calculate RPN in FMEA?
Multiply the three ratings: RPN = Severity × Occurrence × Detection, each rated 1 to 10. A failure mode rated S = 7, O = 5, D = 7 has an RPN of 245. The result ranges from 1 to 1,000, but only 120 different values are possible, because many products cannot be made from three numbers between 1 and 10.
What is a good or acceptable RPN?
There is no universal acceptable RPN. Many companies used a cut-off such as 100, but the 2019 AIAG & VDA handbook advises against RPN thresholds because the same RPN can describe very different risks. On the AP table, a cut-off of 100 would miss 21 S/O/D combinations rated High and flag 55 rated Low.
What is Action Priority (AP) in FMEA?
Action Priority is the method in the 2019 AIAG & VDA FMEA handbook that replaces RPN for deciding which failure modes need action. A fixed table assigns each of the 1,000 S/O/D combinations High, Medium or Low, weighting severity first, then occurrence, then detection. It ranks the need for action, not the risk itself.
What do High, Medium and Low Action Priority mean?
High: the team must either act to improve prevention or detection, or justify and document why current controls are enough. Medium: the team should act, or the company may accept a documented justification. Low: the team could act. The handbook also recommends that management review severity 9–10 items rated High or Medium.
Is severity 10 always High priority?
No. With severity 9 or 10, any occurrence of 6 or more is High, but an occurrence of 1 (the cause cannot happen) is Low, and occurrence 2–3 with strong detection (D 2–4 or 1) is also Low. Severity still dominates: no combination with severity 9–10 and occurrence 4 or more is Low.
Is RPN still used?
Yes. Many companies, customers and older standards still ask for RPN, and plenty of medical-device and non-automotive FMEAs use it. Automotive suppliers working to AIAG & VDA use Action Priority. This calculator shows both, so you can report whichever your customer asks for.
What is the difference between DFMEA and PFMEA?
A design FMEA (DFMEA) looks for ways the product design can fail to meet its function, for example a bracket that cracks under load. A process FMEA (PFMEA) looks for ways the manufacturing or assembly process can make a bad part, for example a bearing pressed in tilted. Both use the same S, O, D ratings and the same AP table, with different rating criteria.
Related on Skill-Lync
- FMEA guide: DFMEA vs PFMEA, the 7 steps and a worked example
- DFM checklist with sheet metal, moulding and CNC rule calculators
- Tolerance stack-up calculator: worst case, RSS and Monte Carlo
- GD&T symbols explained: chart, datums and MMC
- Design for manufacturing and assembly (DFMA)
Sources
- AIAG & VDA, Failure Mode and Effects Analysis – FMEA Handbook, 1st edition, 2019 (AP method, the High/Medium/Low definitions and the management-review recommendation for severity 9–10).
- Relyence, FMEA AP (Action Priority): public reproduction of the DFMEA/PFMEA AP table used to check every cell.
- Schematex FMEA engine (npm schematex 1.1.6), actionPriority(): an independent open-source implementation; it agrees with this page on all 1,000 combinations.
- "Action Priority in new FMEA as factor for Resources Management in Risk Reduction", International Journal of Scientific & Engineering Research 12(4), April 2021, eprints.uklo.edu.mk: reproduces the table and the AP definitions.