Most FMEAs are written to satisfy a customer requirement, not to prevent failures. This guide covers how to run an FMEA that actually identifies your highest-risk failure modes — and what changed in the AIAG-VDA 2019 format.
FMEA (Failure Mode and Effects Analysis) is a structured risk analysis method that rates each potential failure mode on Severity (S), Occurrence (O), and Detection (D) — each on a 1–10 scale — to calculate a Risk Priority Number (S × O × D = RPN) and prioritize corrective actions. The 2019 AIAG-VDA harmonized format replaced RPN with an Action Priority (AP) rating of High, Medium, or Low, which most automotive customers now expect. Getting your FMEA right is what separates a quality system that prevents defects from one that just documents them.
Failure Mode and Effects Analysis is one of the most powerful tools in a quality engineer's toolkit when done correctly — and one of the most expensive compliance exercises when done wrong. The difference is in how you run the analysis session. A FMEA written by a single engineer with a spreadsheet template is almost certainly not an effective FMEA. A FMEA developed by a cross-functional team with real process knowledge, using the part and its failure history as inputs, produces analysis that actually prevents failures.
Design FMEA (DFMEA) analyzes potential failure modes in a product design — how the product could fail to meet its design intent, what the effects of those failures would be on the end user, and what design controls (features, tolerances, materials) prevent them.
Process FMEA (PFMEA) analyzes potential failure modes in a manufacturing process — how each process step could fail to produce the required product characteristic, what the effects would be, and what process controls prevent or detect the failure.
Manufacturing quality engineers primarily work with PFMEAs. DFMEAs are typically the responsibility of product design engineering, though quality engineering participates in design review and often facilitates DFMEA development.
The practical boundary: if the failure mode is in the design (the wrong geometry, the wrong material), it belongs in the DFMEA. If the failure mode is in the manufacturing process (the wrong torque was applied, the wrong material was loaded), it belongs in the PFMEA.
In 2019, AIAG and VDA (Verband der Automobilindustrie) jointly published a harmonized FMEA methodology that made significant changes to the traditional AIAG format. If your customers include major automotive OEMs, you are likely expected to use this format.
Key changes in the AIAG-VDA format:
Seven-step methodology: The 2019 format introduces a seven-step process — Planning and Preparation, Structure Analysis, Function Analysis, Failure Analysis, Risk Analysis, Optimization, and Results Documentation. This replaces the less structured approach of earlier formats.
Separate severity, occurrence, and detection scales: These three dimensions of risk are still present, but the rating tables were revised to reflect current industry expectations.
Action Priority replaces RPN as the primary risk indicator. This is the most significant change for engineers familiar with the previous format.
In the traditional AIAG FMEA format, Risk Priority Number (RPN) was calculated as Severity × Occurrence × Detection, producing a number from 1 to 1000. Teams would prioritize failure modes with high RPNs for corrective action.
The problem with RPN: it treats severity, occurrence, and detection as mathematically equivalent, which they are not. A failure mode with Severity 10 (safety hazard), Occurrence 1 (virtually impossible), and Detection 1 (certain detection) produces RPN 10. A failure mode with Severity 2, Occurrence 5, Detection 1 also produces RPN 10. But the first failure mode is vastly more important — a safety hazard that is unlikely and always detected is still a safety hazard.
The AIAG-VDA 2019 format addresses this with Action Priority (AP), which categorizes failure modes as High, Medium, or Low priority based on a lookup table that gives disproportionate weight to severity. A Severity 9 or 10 failure mode is High priority regardless of occurrence and detection ratings.
Organizations transitioning from traditional RPN to AP must recalibrate their processes. Some failure modes that had low RPNs will now be High priority due to high severity ratings. Some high-RPN failure modes will drop to Medium priority.
The most common error in PFMEA development is conflating failure modes with effects or causes.
A failure mode is the specific way a process step fails to produce the required characteristic. It is stated in physical terms describing the deviation from the intended output.
For a torque operation:
The failure mode is the gap between what the process should produce and what it actually produces. Effects are what happens downstream or to the customer. Causes are what in the process produces the failure mode.
Teams that write "operator error" as a failure mode are not writing PFMEAs — they are writing corrective actions without the analysis. Operator error is a cause. The failure mode is what the operator error produces: the under-torqued bolt, the incorrectly oriented component, the missing step.
A PFMEA requires a cross-functional team. At minimum: process engineering (knows the intended process), manufacturing supervision (knows how the process actually runs), quality engineering (knows the control requirements), and an experienced operator or setup technician (knows how the process fails).
Running an effective session requires:
Preparation: The process must be documented before the session begins. The team cannot analyze failure modes in a process that has not been defined. Use your current process flow diagram and control plan as inputs.
Structure the session around the process flow: Analyze each process step in sequence. For each step, the team identifies the intended function, the potential failure modes for that function, the effects of each failure mode, and the causes of each failure mode.
Don't rush the occurrence and detection ratings: Teams consistently underestimate occurrence rates for failure modes that "never" happen — until they do. Use historical nonconformance data, customer complaint records, and supplier data as inputs to occurrence ratings. Use actual current controls, not aspirational controls, for detection ratings.
Ensure the team disagrees: A room that agrees on everything is not doing analysis. A failure mode with no disagreement on ratings should prompt the question: is this rating based on evidence, or are we all just agreeing to get through the exercise?
The control plan is the implementation of the PFMEA's risk mitigation decisions. For every failure mode in the PFMEA that results in a recommended action, that action should appear in the control plan as a control method.
A high-severity failure mode with no current prevention controls should appear in the control plan as 100% inspection with a specific measurement technique. A medium-severity failure mode with a capable prevention control (error-proofing device) might appear as periodic monitoring.
The FMEA and control plan must be consistent. An auditor traces a high-severity failure mode from the FMEA to the control plan to verify it is being controlled. If the FMEA shows a severity-9 failure mode with Occurrence 3 and Detection 8 — meaning the failure could happen and might not be caught — but the control plan shows no special control for that characteristic, there is a finding.
FMEAs are living documents. They must be reviewed when:
The trigger for review is not the anniversary of the last review. It is any event that changes the risk profile of the process. A FMEA that has not changed in three years in a dynamic manufacturing environment is not being maintained.
Coplain links your FMEA, control plan, and work instructions into a consistent quality control framework. Try free at coplain.com.
Q: What is the difference between DFMEA and PFMEA?
A: DFMEA (Design FMEA) analyzes failure modes in the product design — how the design could fail to meet its intended function. PFMEA (Process FMEA) analyzes failure modes in the manufacturing process — how each process step could produce a nonconforming part. Manufacturing quality engineers primarily work with PFMEA; DFMEA is primarily a design engineering responsibility.
Q: How is RPN calculated in FMEA?
A: RPN (Risk Priority Number) = Severity × Occurrence × Detection, with each factor rated 1–10. A Severity of 9 (risk of injury without warning), Occurrence of 6 (moderate probability), and Detection of 7 (detection unlikely) gives RPN = 9 × 6 × 7 = 378. Under the 2019 AIAG-VDA format, Action Priority (High/Medium/Low) replaces RPN as the primary prioritization tool.
Q: When should FMEA be updated?
A: FMEA should be updated when: a new design or process is introduced, an engineering change occurs, a customer complaint or field failure reveals an undetected failure mode, a new process risk is identified, or a periodic review interval (typically annual) is reached. A FMEA unchanged for more than 24 months in a dynamic manufacturing environment is likely not reflecting current risk.
Q: What is the AIAG-VDA FMEA format and why does it matter?
A: The AIAG-VDA FMEA Handbook (2019) is a joint publication from the Automotive Industry Action Group (AIAG) and the German automotive standards body (VDA). It harmonized two previously separate FMEA methodologies. Most major automotive OEMs now require the AIAG-VDA format — if your customers include GM, Ford, Stellantis, BMW, or Toyota, you need to be using it.
Q: What Severity rating requires immediate action regardless of RPN?
A: Under the AIAG-VDA Action Priority system, any failure mode with a Severity of 9 or 10 (potential for injury or non-compliance with regulations) automatically receives a High Action Priority regardless of Occurrence or Detection ratings. Severity 9 and 10 cannot be managed by improving detection alone — design or process changes must reduce the risk at the source.
Coplain turns any work instruction into a print-ready, audit-proof job aid in minutes.
Try Coplain free →AS9100 Rev D Documentation Checklist: 12 Items Auditors Check First
Most audit failures aren't about process gaps. They're about documentation that doesn't reflect reality. Here's the checklist we wish existed before our first Rev D audit.
6 min readCOMPLIANCEFDA 21 CFR Part 820 Documentation Requirements: What Trips Up Manufacturers
Warning letters. 483 observations. Consent decrees. Most FDA enforcement actions have one thing in common: inadequate documentation. Here's what the agency actually looks for.
10 min readCOMPLIANCEISO 9001:2015 Documentation Guide: What You Actually Need vs What You Think
The 2015 revision eliminated the quality manual requirement and six mandatory procedures. What it added is more demanding. Here's what actually changed — and what your registrar looks for during surveillance.
7 min readRelated Tools & Templates
COPLAIN PLATFORM
See every tool that turns complex work instructions into floor-ready documents.