What is the difference between RPN and Action Priority in FMEA?

RPN multiplies severity, occurrence, and detection into a number from 1 to 1000, treating all three factors equally. Action Priority (AP) looks the same three scores up in a table that returns High, Medium, or Low and weights severity first. RPN gives a fine but flat rank; AP gives a coarser priority that will not let a severe failure be averaged away.

Why did AIAG-VDA replace RPN with Action Priority?

Because RPN's scale is flat: the same number can come from a rare safety hazard or a frequent cosmetic defect, and a high-severity failure with low occurrence can score lower than a harmless common one. The 2019 AIAG-VDA FMEA handbook replaced the single RPN threshold with an Action Priority table that evaluates severity first, so a safety-critical failure can never be rated Low priority.

Is RPN still valid or completely obsolete?

RPN is legacy but not banned. It is still widely used and is a reasonable quick rank for low-stakes internal exercises, and some customers and older documents still require it. For any FMEA where a severe failure must never be masked, the current AIAG-VDA method uses Action Priority instead.

Does Action Priority use the same severity, occurrence, and detection scores as RPN?

Yes. Both methods use the same three 1-to-10 ratings with the same scoring criteria. Only the final step, how you convert the three scores into a priority, changes. If your team already scores the three ratings consistently, that work carries straight over to AP.

How does the Action Priority table weight severity first?

The table is read in order: severity sets the starting band, occurrence narrows it, and detection acts as the tie-breaker. Because severity is evaluated first and most heavily, a severity 9 or 10 failure stays in High priority regardless of how good its occurrence and detection scores are. Detection can move an item between Medium and Low but cannot pull a safety failure down on its own.

How do you transition an existing FMEA from RPN to AP?

Calibrate the ratings first, then add an Action Priority column next to RPN and run both for a cycle. Investigate every case where the two disagree, especially a low RPN with a High AP, re-sort the action list by AP, and address every High item. Once the team trusts AP, make it the standard output and retire the RPN threshold, keeping RPN only where a customer requires it.

What is a good RPN threshold to act on?

There is no universal number, which is part of why the industry moved away from RPN. Thresholds like 100 or 125 are arbitrary and miss high-severity items with modest RPNs. Action Priority removes the threshold entirely: a fixed table decides High, Medium, or Low, so there is nothing to set. Regardless of method, always act on high-severity failures.

RPN vs. Action Priority (AP) in FMEA

Why AIAG-VDA retired the Risk Priority Number for a table that puts severity first, and how to switch without breaking the FMEAs you already have.

RPN and Action Priority both rank FMEA risk from the same three scores, severity, occurrence, and detection, but they combine them differently. RPN multiplies the three into a number from 1 to 1000; Action Priority (AP) looks the combination up in a table that returns High, Medium, or Low and weights severity first. The 2019 AIAG and VDA FMEA handbook replaced RPN with AP.

This is not a naming change. RPN and AP can rank the same failure modes in a different order, and a team that switches without understanding why can either over-invest in trivial risks or, worse, wave through a safety failure because its number looked small. This guide explains what each method does, the flat-scale problem that pushed the industry off RPN, how the AP table puts severity in charge, and how to transition without breaking the FMEAs you already have.

What is RPN?

The Risk Priority Number is severity times occurrence times detection: RPN = S × O × D. Each factor is scored 1 to 10, so RPN ranges from 1 to 1000. You calculate it for every failure mode on the FMEA worksheet, sort the list high to low, and work the top items first. For decades this was the standard way to prioritize, and its appeal is obvious: one number, easy to sort, easy to explain. Many teams set a threshold, act on everything above 100 or 125, and moved on.

That simplicity is also the problem. RPN treats a point of severity as interchangeable with a point of occurrence or detection, because multiplication does not care which factor is large. An unsafe failure and a cosmetic one can land on the same number, and the number gives you no way to tell them apart.

What is Action Priority (AP)?

Action Priority is a lookup, not a calculation. Instead of multiplying, you take the specific combination of severity, occurrence, and detection and read a priority of High, Medium, or Low from a defined table that covers all 1,000 combinations. The table is built so that severity dominates: a high-severity failure lands in High priority almost regardless of how low its occurrence and detection scores are, while a low-severity failure has to have genuinely bad occurrence and detection to climb out of Low. The output is not "how much risk" as a number; it is "how urgently does this need action."

Because AP is a fixed table published in the handbook, every team that adopts it prioritizes the same combination the same way. Two engineers scoring an identical failure get an identical priority, which removes the argument about where to set an RPN threshold, there is no threshold to argue about.

Both failures multiply to an RPN of 200, so RPN ranks them the same. Action Priority weights severity first, so the safety failure comes out High and the cosmetic one Medium.

Why did AIAG and VDA replace RPN?

Because the RPN scale is flat, and a flat scale hides the failures you most need to see. The number 200 can be built from severity 10, occurrence 5, detection 4, a genuine safety hazard, or from severity 5, occurrence 8, detection 5, a frequent cosmetic nuisance. RPN reports both as 200. Rank purely on RPN and the two sit side by side, and a threshold of, say, 250 would drop both below the action line, quietly filing an unsafe failure alongside a paint blemish.

Two well-known distortions follow. First, thresholds are arbitrary: nothing makes 125 meaningful, and shifting it a little reshuffles the whole action list. Second, and more dangerous, high-severity failures with low occurrence and good detection produce modest RPNs, so a rare-but-catastrophic failure can score lower than a common-but-harmless one. That is exactly backwards for anything involving safety or regulation. When the automotive industry harmonized the American (AIAG) and German (VDA) FMEA methods in 2019, they addressed this head-on by retiring the single RPN number in favor of the Action Priority table, so severity can never be averaged away.

How does the AP table weight severity first?

The table is read in a strict order: severity, then occurrence, then detection. Severity sets the band you start in, occurrence narrows it, and detection makes the final call within that band. Because severity is evaluated first and most heavily, a high-severity failure cannot be rescued into Low priority by good occurrence and detection scores, the structure of the table simply does not allow it. Detection, which multiplication treated as an equal partner, becomes the tie-breaker it should be: it can move you between Medium and Low, but it cannot pull a safety failure down to Low on its own.

Action Priority evaluates the three factors in order. Severity sets the band and detection only breaks ties, so a safety-critical failure stays High no matter how the other two scores land.

Does switching to AP change how you score severity, occurrence, and detection?

No. Both methods use the same three 1-to-10 ratings, and the scoring criteria and calibration discipline are identical. If your team already scores severity, occurrence, and detection consistently, that work carries straight over; only the last column, how you turn the three scores into a priority, changes. This is why the transition is more about mindset than rework: you keep the ratings and swap the ranking. A team with sloppy ratings will get sloppy AP results just as it got sloppy RPNs, so the calibration matters more than the method.

	RPN	Action Priority (AP)
How it combines S, O, D	Multiplies: S × O × D	Looks up the combination in a table
Output	A number, 1 to 1000	High / Medium / Low
Severity weighting	Equal to the other two	Dominant; evaluated first
Action rule	Team-chosen threshold (arbitrary)	Fixed table (no threshold to set)
Main weakness	Flat scale hides severe, rare failures	Coarser: three bands, not a fine rank
Status	Legacy; still widely used	Current AIAG-VDA 2019 method

RPN and AP score the same three factors. They differ only in how they turn those scores into a priority, and in how much they let severity dominate.

How do you transition from RPN to AP?

You do not have to rip up every FMEA overnight. The clean path is to keep your ratings, add the AP result, and let the two run side by side until the team trusts the new priorities.

Confirm the ratings first. AP is only as good as the severity, occurrence, and detection scores behind it. Calibrate the team on the rating criteria before you change anything downstream, because a method change cannot fix bad inputs.
Add an AP column beside RPN. On existing FMEAs, keep the RPN column and add Action Priority from the table. Running both for a cycle lets the team see where the two disagree and understand why.
Investigate the disagreements. Wherever a low RPN carries a High AP, or a high RPN carries a Low AP, stop and look. These are exactly the cases the change was made to catch, usually a severe, rare failure that RPN was underrating.
Re-sort the action list by AP. Prioritize by High, then Medium, then Low, and address every High item regardless of what its old RPN said. Within a band, you can still use the underlying scores to sequence work.
Update the FMEA template and training. Once the team trusts AP, make it the standard output on the form, retire the RPN threshold rule, and train new members on the table rather than on a cutoff number.
Keep both only where a customer requires it. Some customers or legacy documents still ask for RPN. Report it where contractually required, but drive your own actions from AP.

Which should you use?

If you are in automotive or any supply chain that follows the harmonized handbook, AP is the current expectation, and its severity-first logic is the safer default anywhere failures can hurt someone. RPN is not wrong so much as blunt: it is still a reasonable quick rank for low-stakes internal exercises, and it remains common enough that you should know how to read one. But for any FMEA where a severe failure must never be masked, a safety feature, a regulated characteristic, a critical-to-quality dimension, AP is the better tool because it refuses to average severity away. The deeper point holds for both: the ranking method only decides where you look first. Whether you use a number or a table, the FMEA is only honest if the underlying scores reflect how the process actually behaves.

The sources behind RPN and Action Priority

The American Society for Quality documents the classic FMEA approach, scoring severity, occurrence, and detection on 1-to-10 scales and multiplying them into a Risk Priority Number from 1 to 1000 (ASQ, What Is FMEA?).
The AIAG and VDA FMEA Handbook, first published in 2019, harmonized the American and German automotive FMEA methods and replaced the single RPN threshold with an Action Priority (High / Medium / Low) table that evaluates severity first (AIAG, FMEA).
The Action Priority table assigns a priority to every one of the 1,000 severity-occurrence-detection combinations, so a safety-critical failure with severity 9 or 10 cannot be rated Low priority regardless of its occurrence and detection scores (AIAG-VDA FMEA).

How does a connected floor keep either method honest?

Both RPN and AP rest on the same foundation: occurrence and detection scores that are supposed to describe how the process behaves right now. When defect counts, escapes, and detection results live on clipboards and month-end tallies, teams re-score from memory, and a failure mode that is trending worse in the real world keeps its old, comfortable rating on the worksheet. Neither multiplying nor looking up in a table can fix a stale input. When that data is captured live at the point of work, every defect, every escape, tied to the line and the cause, the ratings can be re-scored against reality, and a rising occurrence surfaces before it becomes a complaint. That live feedback is what Harmony gives a plant, working alongside the tools you already use, no rip-and-replace. The processor in our CLS case study moved from failure data found the next morning to failure data visible during the shift, which is the loop any FMEA needs, whichever way it ranks risk. Related reading: the AIAG-VDA FMEA method in full, and how the same scores drive process FMEA and design FMEA. See the capture on the features overview.