Quality Engineering: Origins & Structure of PFMEA

In the first part of our Quality Engineering blog series, we discussed the potential to design a manufacturing process that was robust against human errors. The Process Failure Mode and Effects Analysis (PFMEA) is the tool used by the automotive industry (and others) to try to make this a probability.

Competition in the auto industry is intense, and as a result, the daily battles between cost and quality are cutthroat in the design and manufacture of vehicles, where saving a single cent on a part and preventing even one warranty defect are considered major victories. Conspiracy theories like notorious “Detroit has a car that runs on water” one are ludicrous; engineers exhaust themselves trying to get a half-mile per gallon improvement from a vehicle, for instance. Automakers go to extraordinary lengths to win these battles and they invent/adopt, use, and refine improvement tools that are used along the tiered supply chain to help.

Origins & Structure of PFMEA

Origins

The PFMEA in its current state has been around for about 25 years. Its origins are in the FMECA (Failure Mode Effect and Criticality Analysis) developed by the US military for analyzing defense systems in the late 1940’s. It was a design-oriented analysis and it was heavily adopted by the civil aviation industry in the late 1950’s. Automotive companies adopted it as the Design Failure Modes and Effects Analysis (DFMEA) in the 1970’s after some scary product problems (like with the Pinto). The FMEA concept has been applied to Design, Concept, System, Tooling, Machinery, and Process methods (in the early 1950s) among others.

We’ll discuss the Process FMEA (PFMEA) since we’re talking about your manufacturing process that still has error-prone humans involved. The PFMEA has been tweaked, polished, expanded, and re-applied over the last 25 years, but its purpose and goal have not changed: the PFMEA is an analytical tool to review the steps in a manufacturing process for quality risks.

The PFMEA doesn’t determine if you have a good process or a bad one; it tells you nothing about your process’s efficiency or complexity. It also doesn’t show how to improve your process (directly). Instead, PFMEA helps you identify what could go wrong with what you’re planning or what you already have. Further, it identifies where you need to concentrate and put your priorities. Since PFMEA is only an analysis tool, mitigating risk is up to your team.

It’s vital to understand though, that the PFMEA isn’t used once and then put away; it’ can be used at every iteration of process development. For example, if a colleague has an idea, analyze it and then update the PFMEA with it. Or maybe you make a change to an existing process. Analyze it with PFMEA. This is especially important during the first steps designing a new process, when you haven’t incurred any equipment costs yet. Use a version of Deming’s PDCA Loop (Plan/Do/Check/Act).

• Plan: “What if we did this?”
• Analyze: “Pull out the PFMEA and analyze it”
• Evaluate: “Look at the quality risks we uncovered.”
• Revise: “What countermeasures could we put in place against these risks?”

Structure

The PFMEA is composed of two major sections: the header and the main body. An example of the header can be seen below in Figure A:

Figure A

The header lists the scope (partnumber, program, and customer), approvals, revision dates, and the team leader. It’s there to help with revision control, mostly.

The main body (see Figure B) is a table of around 20 columns (this is the AIAG version):

Figure B

The rows in the table are the process steps in the manufacturing process, one or more for each process step. Below is a list of the columns as they would appear in the table from left to right, and what they’re used for:

• Process Step No. — a number to keep track of the process steps (not shown in Figure B)
• Process Step Description — what the process step does (e.g. “insert screw”)
• Process Function Requirement — what a successful process step looks like (describes success)
• Potential Failure Mode — how the process can be unsuccessful
• Potential Effects of Failure — how the failure can manifest itself in the product
• Severity Rating (S) — how bad the effect is (from a table 1–10)
• Classification — reserved for special designations, internal or customer
• Potential Causes of Failure — why the process step was unsuccessful
• Prevention Controls — how we try to make people aware of the potential failure mode
• Occurrence rating (O) — how often does the failure mode occur (frequency) (from a table 1–10)
• Detection Controls — these range from not allowing the failure mode to occur to detecting it
• Detection Rating (D) — the strength of the detection control (from a table 1–10)
• RPN (Risk Priority Number) — the product of Severity, Occurrence and Detection RPN = S x O x D (1 to 1000)
• Recommended Action — recommended countermeasures to reduce the RPN, usually via O or D
• Responsibility — who undertakes the search for a countermeasure
• Target Completion Date — when
• Actions Taken — a description of the actual actions taken as countermeasure
• Revised Severity Rating — the new Severity rating (it is unusual to change the Severity rating in a PFMEA)
• Revised Occurrence Rating — the new Occurrence rating
• Revised Detection Rating — the new Detection rating
• Revised RPN — the new product of S x O x D

Every row of the PFMEA has these columns. Some column entries repeat row-to-row as you analyze a particular process step; it may have multiple process function requirements, multiple failure modes, multiple effects or multiple controls. The length shouldn’t scare you; the deeper, more exhaustive, and more meticulous the analysis, the better chance your team has of stopping a problem from being made and/or getting out the door.

Conclusion

That’s the PFMEA. It looks both simple and complicated; a simple structure with a lot of depth. And will be as long as your process is. PFMEA is especially valuable during the planning stage — before any equipment or process is designed or purchased. Couple this type of analysis with your supply chain knowledge to help your organization grow, faster, better, and more profitably.

In the next part of our Quality Engineering blog series, we’ll examine how PFMEA’s are misused and what makes them loathed by many professionals, across industries. Thanks for reading.