New FMEA manual offers more flexibility
by R. Dan Reid
The new fourth edition of the Chrysler, Ford and General Motors (GM) Potential Failure Mode and Effects Analysis (FMEA) Reference Manual,1 which was released last year, is a significant upgrade from the third edition published in 2001. It is available from the Automotive Industry Action Group (AIAG).
ASQ defines FMEA as “a systematized group of activities to recognize and evaluate the potential failure of a product or process and its effects, identify actions that could eliminate or reduce the occurrence of the potential failure and document the process.”2
The new edition itself says FMEA is “an analytical methodology used to ensure that potential problems have been considered and addressed throughout the product and process development process (APQP—advanced product quality planning).”3
FMEA is an effective tool for quantifying risk so it can be analyzed, prioritized, mitigated and eliminated. FMEA can be applied in all types of organizations—from manufacturing to service. To be optimal, FMEA should be performed by a cross-functional team during product or process development as part of the APQP process.
Chrysler, Ford and GM first published the common FMEA reference manual in 1993 as a result of discussion during a 1988 ASQ Automotive Division conference. Suppliers at that conference reported to a panel of the original equipment manufacturing (OEM) purchasing vice presidents that there was money to be saved by harmonizing the various company-specific requirements existing at that time.
Subsequent supplier feedback indicated FMEA was one of the priority topics to work on, so committees were formed under the auspices of AIAG to begin work on this and a number of other projects, such as APQP, measurement systems analysis, production part approval and statistical process control.4
The two types of FMEA—design and process—are addressed in the new edition, along with three possible scopes—system, subsystem or component. There is more flexibility in creating and assessing risk with FMEA in this edition. Appendix A contains new optional forms: six for design FMEA and eight for process FMEA.
Users can choose forms that highlight certain relationships between the data—such as prevention controls to occurrence ranking—to facilitate sorting by date, to assist in determination of failure modes or to provide consistency between the process flow chart, the process FMEA and control plan. These replace one standard form for each type in the previous editions.
Earlier editions quantified risk solely by use of a risk priority number (RPN) calculated by multiplying ratings for severity (S) of problem by likelihood of occurrence (O), by likelihood of detection (D), or S x O x D = RPN.
In the years since the initial release of the FMEA reference manual, some automotive customers strongly emphasized supplier RPN reduction programs. In some cases, suppliers were mandated to work to reduce all RPNs under a certain value (40, for example).
Use of such arbitrary thresholds is not recommended in the new edition. Organizations should be working on the highest priority risks regardless of the rating criteria or values. High severity ratings should always be addressed.
Appendix C in the new edition offers other alternative risk analysis criteria for use with the FMEA. One interesting alternative is called SOD or SD. It is a non-arithmetic combination of the ratings shown earlier. For example, for severity = 6, occurrence = 3 and detection = 7, the SOD rating would be 637. The SD calculation is the same, excluding the occurrence value. So the SD rating for the previous example would be 67. The various SOD ratings would then be sorted in numerical descending order to get the highest severity scenarios at the top of the list.
Appendix D in the fourth edition lists some alternative techniques to FMEA for analyzing risk. These include failure mode, effect and criticality analysis, design review based on failure modes, and fault tree analysis. Examples of some of these are provided in the appendix. These techniques can complement the use of FMEA, but their use as a replacement for FMEA may be subject to customer approval.
Tips For Effective FMEAs
A good failure mode effects analysis (FMEA) should:
completed across to the right side of the FMEA form, which is the area for
recomputing the risk after initial efforts have been taken to reduce the initial high risk rating in some cases—for example, risk priority number (RPN).
- List multiple effects for each mode and multiple causes for each effect. Generally, there is not a 1:1 relationship because any given failure mode could have many effects.
- List actions taken on high severity and high RPN or other risk ratings, aimed at preventing the occurrence of a potential failure. Responsibility for the actions taken should be assigned and tracked to completion.
- Include efforts to error-proof the design and process.
- Ensure that measurement uncertainty is known and adequate for applicable metrics.
- Identify characteristics that should be designated as special or critical on the control plan so actions can be planned and implemented to mitigate the effects of the potential failure. This should include reference to a contingency plan to protect your customer from receiving noncompliant product.
- Carefully consider the risk (for example, safety, quality, equipment and resource) and the efficiency of methods used, actions taken and contingency planning.
- Consider the major types of design failures (materials, processes, costs) or the major types of process failures (too much, too little, missing or wrong).
The fourth edition indicates that the scope establishes the boundary of the FMEA analysis.5 The scope can be system, subsystem or component (see Figure 1). A system consists of many subsystems, which consists of many components. The links and interactions among parts within the scope and with other subsystems or systems also have to be addressed.
The revision lists a number of tools that can help define the appropriate scope, which should dictate the appropriate team membership. These tools include block diagrams, process flow charts, schematics and bills of material.
The suggested evaluation criteria for S, O and D ratings provided in the new edition have been enhanced to be more meaningful to real world analysis and use.6 In some cases, they provide more consideration for the effects on the vehicle. They continue to provide 10 ranking options per table, which provides adequate discrimination for prioritization in the final calculations.
In addition to an index and list of helpful references, the revision, like previous editions, has numerous examples and diagrams to provide improved guidance. There is a new section on linking FMEAs to other documents, including the block diagram and the design verification plan and report.
Using the new benchmark FMEA process with these tips should ensure an effective risk mitigation process in your organization.
References and note
- Chrysler LLC, Ford Motor Co. and General Motors, Potential Failure Mode and Effects Analysis (FMEA) Reference Manual, fourth edition, AIAG, 2008.
- “Quality Glossary,” Quality Progress, June 2007, p. 46.
- Chrysler LLC, Ford Motor Co. and General Motors, see reference 1, p. 2.
- See R. Dan Reid, “TS-16949–Where Did It Come From,” Quality Progress, March 2005, pg. 32, for more background information.
- Chrysler LLC, Ford Motor Co. and General Motors, see reference 1, p. 10.
- Chrysler LLC, Ford Motor Co. and General Motors, see reference 1, p. i.
R. Dan Reid, an ASQ fellow and certified quality engineer, is a purchasing manager at General Motors Powertrain. He is co-author of ISO/TS 16949; the Chrysler, Ford, GM APQP, PPAP and Potential FMEA manuals; ISO 9001:2000; ISO IWA 1; and AIAG’s new Business Operating Systems for Health Care Organizations.