DFMEA (Design Failure Mode and Effects Analysis) finds ways a product's design could fail before it goes into production. PFMEA (Process Failure Mode and Effects Analysis) finds ways the manufacturing process could fail to build that design correctly. DFMEA asks "could this design fail?" PFMEA asks "could we make it wrong?"Most manufacturers need both: DFMEA during development, and PFMEA from process planning onward.
Key Points
- DFMEA is owned by design engineering and completed before the design is released. PFMEA is owned by manufacturing or process engineering and maintained for as long as the process runs.
- DFMEA outputs feed PFMEA. Special characteristics identified in the design become controls in the process.
- The AIAG-VDA FMEA Handbook (2019) replaced the traditional Risk Priority Number (RPN) with Action Priority (AP) for ranking which risks to act on first.
- A PFMEA is only as accurate as its Occurrence and Detection ratings. When those are estimated once and never updated, the FMEA drifts away from what's actually happening on the line.
What is FMEA?
Failure Mode and Effects Analysis (FMEA) is a structured approach used to identify and prioritize potential failure modes within a system, product, or process. It systematically examines how and where failures might occur and assesses their potential impact. By highlighting these vulnerabilities early, teams can take action to mitigate risks, improve reliability, and reduce the chance of costly disruptions or defects.
FMEA evaluates three critical factors for each failure mode:
- Severity - How serious would the consequences be if the failure occurred?
- Occurrence - How likely is the failure to happen?
- Detection - How easily can the failure be detected before it causes an issue?
Traditional FMEAs multiply these ratings into a Risk Priority Number (RPN). The AIAG-VDA FMEA Handbook, now the standard across much of the automotive supply chain, uses Action Priority (AP) instead. AP ranks each risk High, Medium, or Low and weights Severity most heavily, so a high-severity failure can't hide behind a low RPN.
FMEA has two main variations: Design FMEA (DFMEA) and Process FMEA (PFMEA)-each serving distinct but equally critical roles in preventing failures during different stages of production.
What is DFMEA?
Best for new or changed product designs. Finds failure modes caused by design decisions, before tooling and production begin.
Design Failure Mode and Effects Analysis (DFMEA) is a risk assessment tool focused on identifying potential failure modes during the product design phase. Its primary goal is to anticipate how a product's design could fail, addressing these issues before they reach manufacturing or, worse, the customer.
- Owner: Design or product engineering, with input from manufacturing, quality, and suppliers
- When: Early concept through design release, and again whenever the design changes
- Typical inputs: Design requirements, boundary diagrams, P-diagrams, warranty and field-failure history
- Typical outputs: Design changes, test plans (such as the DVP&R), and special characteristics passed to PFMEA
- Example: A brake caliper piston seal made from a compound that hardens under repeated high-temperature cycles, which could lead to fluid leaks and reduced braking
What is PFMEA?
Best for new or changed manufacturing processes. Finds failure modes caused by process steps, equipment, materials, or methods.
Process Failure Mode and Effects Analysis (PFMEA) is a method used to evaluate risks and potential failure modes within the manufacturing process. Its goal is to identify and mitigate problems that could lead to inefficiencies, defects, or equipment failures during production.
- Owner: Manufacturing, process, or quality engineering
- When: During process planning, before production launch, and throughout production whenever the process, equipment, or materials change
- Typical inputs: Process flow diagram, DFMEA special characteristics, equipment capability data, scrap and rework history
- Typical outputs: Process changes, error-proofing, and the control plan
- Example: A press-fit station applying too little force, leaving the same caliper seal partially seated
DFMEA vs PFMEA: Key Differences
| DFMEA | X | PFMEA |
|---|---|---|
| DFMEA examines potential design failures that could affect product function, performance, and user safety. It addresses how flaws in the design itself can lead to product malfunction or failure in real-world applications. | Scope and Focus |
PFMEA targets risks within the manufacturing process, focusing on identifying inefficiencies or process breakdowns that could lead to defects, waste, or production delays. |
| DFMEA is applied during the product development or redesign phases, ensuring that the product is designed to meet all safety, performance, and reliability standards before it enters production. | When They Are Used |
PFMEA is typically used before or during full-scale production to ensure that the manufacturing process is stable, efficient, and capable of producing high-quality products consistently. |
| DFMEA involves collaboration between product design teams, engineers, and quality specialists. Their focus is on refining the design to prevent future product issues. | Teams Involved |
PFMEA engages manufacturing teams, including production supervisors, process engineers, and quality control experts, who work together to optimize the production line and address potential process-related risks. |
| DFMEA results in design improvements, such as changes to materials, structural adjustments, or enhanced features that improve the product's overall safety and performance. | Outcomes | PFMEA leads to process optimizations, including improved equipment maintenance, operator training, workflow modifications, or adjustments in environmental conditions to ensure a smoother, defect-free production process. |
How DFMEA and PFMEA Work Together
DFMEA and PFMEA are linked, not separate exercises. When the DFMEA flags a characteristic as critical (the seal's seating depth, for example), that characteristic moves into the PFMEA. The PFMEA then identifies every process step that could affect it and sets controls for each one. Those controls go into the control plan, which tells operators what to check, how often, and what to do when something falls out of spec.
Information also flows the other way. If the PFMEA shows a characteristic can't be held reliably with the current process, that's a reason to revisit the design.
Industries with high failure costs rely on this link most. Automotive suppliers working to IATF 16949 are expected to maintain both, and aerospace and medical device manufacturers use the same design-to-process link to meet their own quality and regulatory requirements.
Why PFMEAs go stale
Most PFMEAs are built in a cross-functional workshop before launch. The team estimates how often each cause will occur and how well current controls will catch it. Then the document gets approved, filed, and reopened when something goes wrong.
The line doesn't stand still in the meantime. The equipment running the process wears. A spindle bearing starts to degrade, a hydraulic pump loses pressure, a press drifts out of its force window. Each change raises the real Occurrence of a failure cause the PFMEA still rates as unlikely. Usually nobody knows until scrap goes up, a defect reaches a customer, or an audit asks why the ratings haven't changed in three years.
The costs are familiar: rework and scrap that eat into first-pass yield, repeat failures handled as one-offs, containment actions that pull engineers off improvement work, and, in the worst cases, recalls or regulatory findings.
How to keep PFMEA accurate with equipment condition data
A PFMEA stays useful when its ratings reflect what the equipment is actually doing. That takes three capabilities:
- Continuous condition data on process-critical equipment. Vibration, ultrasound, temperature, and electrical monitoring on the machines tied to special characteristics show when a failure cause is developing, not just after it has produced defects. That's real evidence for Occurrence ratings, and continuous monitoring can serve as a Detection control for equipment-driven failure modes.
- A closed loop from alert to verified fix. An alert only lowers risk if it becomes action. The alert should create a work order, the repair should be verified with before-and-after data, and the outcome should be recorded, so the PFMEA reflects the fix that actually worked.
- Integration with the systems teams already use. Condition data belongs in the existing CMMS and quality workflows, not a separate screen someone has to remember to check.
Tractian is built around these three capabilities. Its sensors continuously monitor the health of rotating and electrical equipment, and AI diagnostics identify the developing failure mode and its severity. Tractian also integrates with leading CMMS platforms, so alerts become work orders in the system your team already runs. Every intervention is then tracked through to a verified, healthy asset, which gives reliability and quality teams traceable evidence to update their FMEAs with.
At Cadillac Casting, a 280,000 sq ft foundry, early detection of an elevator bearing failure meant the team could run a root cause analysis with the bearing supplier on-site, instead of replacing the same part again and again. When Tractian flagged a lubrication issue on a drag shakeout motor at go-live, a supervisor greased it right away and watched its health status return to green on the platform, confirming the fix.
At Pirelli's tire plant, technicians checked in on 98% of alerts, the team identified 77 failures, and the monitored exhaust systems have had zero recorded breakdowns since deployment.
FAQs about DFMEA and PFMEA
Which comes first, DFMEA or PFMEA?
DFMEA usually starts first, because the process can't be analyzed until the design is defined. In practice the two overlap. PFMEA work often begins before the design is frozen so manufacturability issues can shape it.
What does PFMEA stand for?
PFMEA stands for Process Failure Mode and Effects Analysis. It's sometimes written P-FMEA or Process FMEA.
Who is responsible for PFMEA?
Manufacturing or process engineering typically leads the PFMEA, with quality, maintenance, operators, and design engineering contributing. Maintenance input matters because equipment condition drives many process failure causes.
What is the difference between RPN and Action Priority?
RPN multiplies Severity × Occurrence × Detection into one number, so very different risks can end up with the same score. Action Priority, introduced in the AIAG-VDA FMEA Handbook, ranks each combination as High, Medium, or Low and gives the most weight to Severity.
How does PFMEA relate to the control plan?
The PFMEA identifies what could go wrong in the process and which controls address it. The control plan turns those controls into day-to-day instructions: what to measure, how often, and how to react.
How often should a PFMEA be updated?
Update the PFMEA whenever the process, equipment, materials, or design changes, and whenever real-world data (scrap, defects, equipment condition) shows that a rating no longer matches reality.