The FMEA process was developed in the late 1940s and for 70 years, there was little change in FMEA methodology. However, in 2019, AIAG (the US Automotive Industry Action Group) and their German counterpart, VDA, issued substantial changes in the way FMEAs are to be done in the automotive industry. The changes have standardized the FMEA process throughout the global automotive industry and have revised how FMEAs are conducted in a few key areas. Many industries will continue to use the traditional methodology, while those in the automotive industry will immediately begin using the new methodology.
- All FMEAs have the same objective – to reduce risk.
- The process used to identify and reduce risk is similar for both traditional and AIAG-VDA FMEAs.
- Potential failures are evaluated on three scales or evaluation criteria: How severe the failure would be, how likely it is the failure will occur, and how difficult it will be to detect the failure before or after it occurs. These criteria should be customized to be meaningful to the organization using them.
- The failures are then prioritized to reduce the risk.
- FMEAs are done by teams; not individuals.
- FMEAs become living documents that get updated when there are major changes to the design or process.
- FMEAs should be used as the basis for developing a Control Plan. Control Plans are a summary of defect prevention, reactive detection techniques including formal Reaction Plans.
Traditional (RPN) FMEAs
We call the original FMEA methodology “Traditional FMEA (RPNs).” Traditional FMEAs establish an RPN or Risk Priority Number for each failure mode and its resulting effect(s). The RPN is a function of three factors: The severity of the effect, the frequency of occurrence of the cause of the failure, and the ability to detect (or prevent) the failure or effect.
RPN = Severity rating X Occurrence rating X Detection rating; The RPN can range from a low of 1 to a high of 1,000
Once the RPNs are determined, an Action Plan is developed to reduce the risks of failure modes of unacceptably high RPNs.
We refer to the new AIAG-VDA FMEA methodology as “AIAG-VDA FMEAs.” Some of the improvements this methodology makes to the FMEA process are:
- The seven clearly defined steps in the FMEA process are broken into three phases: System Analysis, Failure Analysis and Risk Mitigation, and Communication. While our training in traditional FMEAs is broken into 10 distinct steps that encourage the FMEA team to take a step-wise approach to the FMEA, the generally accepted approach to conducting a traditional FMEA often became an exercise driven by filling out the spaces in the FMEA Analysis Worksheet.
- The concept of the Failure Chain encourages FMEA teams to think about the failure as part of a system.
- Greater specificity for the Severity, Occurrence and Detection Evaluation Criteria assures more accurate evaluation of the failure mode.
- The RPN was eliminated and replaced with Action Priority Tables. Critics of the RPN rightfully were concerned that it could lead teams to focus on reducing the wrong risks. Further, too often teams would use an arbitrary number as a “cut-off” to determine which failures needed to be addressed and which don’t. In fact, because the ratings and therefore the RPN themselves are somewhat arbitrary and also relative to one another, a “cut-off number” isn’t necessarily relevant or reflective of which failures need to be addressed. Critical failures could be missed and minor failures that are addressed could result in wasted time and money.
Which FMEA Methodology Should You Use?
If your customer specifies that you use one or the other process, then that is the one you should use. If not and your business is new to FMEAs, it makes sense to start with the AIAG-VDA method because it provides a much better defined step-by-step process and ultimately will lead to more effective risk reduction. If your business has been using traditional FMEAs and they have been effective, continue using them. We have seen substantial improvements made using traditional FMEAs and it is hard to argue that a process with 70 years of history is ineffective.