AIAG Releases FMEA Update - 30 June 2001

Hello everybody
There is a new AIAG FMEA revision, I haven't read it yet but I just want you to be aware

------------------
Alex

Cha-Ching$$$$$$$$$$$$$$$$$$

Continuous Improvement or Continuous Gouging!

Hello again,

I reviewed the link that Marc posted and the AIAG hasn't released the 3rd edition yet, so I probably have a draft document. I found several changes.

1. CURRENT CONTROLS. Now we have to establish if the controls are for detection or for prevention.

2. SEVERITY CRITERIA. There are two columns, one for the end user effect and the other one for Manufacturing / Assembly Effect.

3. OCCURRENCE CRITERIA. The Cpk column was deleted, but a Ppk matrix with calculation explanation was added in the appendix.

4. DETECTION CRITERIA. It is established a more clear criteria focused in error proofing.

5. RECOMENDED ACTIONS. More clear criteria specifying that rankings of 9 or 8 should have special atencion. And what are the specified steps for recomended actions, first severity, then occurence and finally detection.

My point of view is that this FMEA 3rd edition is very similar of the FMEA manuals released by Ford.

Regards,

------------------
Alex

[Marc]

In the pdf_files directory there is a gif file named FMEA-Newest_Manual_Released.gif announcing the release of the AIAG's 'newest' (Edition 3) FMEA Manual.

I have also received the 'Changes Summary' which I'll be posting in the Members File Access directory as soon as I get them typed in - maybe in a day or three. Of course the Premium Access folks will auto-magically get the summary at the same time as Premium Access includes access to all the files in the Members Access directories.

Stay tuned. Changes are mostly clarifications which more specifically define what they ('Le Grande Menage-a-Trois') 'expect' and/or 'require'. One might even look at most of the AIAG's revisions as one sees the QS-9000 'Interpretations' which periodically auto-magically appear.

This is really only significant to 'Big Menage-a-Trois' suppliers. If you do FMEAs of whatever kind/type and you're not 'automotive', this isn't much worth your attention.

I have not yet contrasted the SAE's FMEA Manual - J1739_200006 (available from the Society of Automotive Engineers ) with the new AIAG FMEA 'Manual' (I guess I'll have to buy it....) so I won't yet comment on how they contrast. I'll do that and publish it when I upgrade the FMEA Guide File Set.

Hi Marc:
Long time no "see". I requested a copy of the changes from AIAG (attached as Word document). Here also is a text version of the 3rd edition changes. My main gripe as an auditor has been addressing not only high RPNs but also those with high severity which the 3rd edition now more clearly recognizes (yes, there is a God):

3rd Edition
Design FMEA – Introduction
· In the Team Effort section, the technical/physical limits of product maintenance (service) is given consideration.
Design FMEA – Development of a Design FMEA
· Detailed discussion of system, subsystem, and component in item 2 with an associated appendix (Appendix F) that provides descriptions and examples.
· For the severity ranking, the ranking used is the most serious for a given failure mode. If a failure mode has numerous effects, only the most serious matters.
· Two notes added to discussion of severity ranking.· Additional examples of typical causes of failure provided.
· Occurrence table changed to failures per thousand vehicles/items.
· Only two types of design controls are recognized:§ Prevention§ Detection
· For design controls, the form has been changed to provide for separate columns to record controls. One column is to be used to record prevention controls, the other column is for detection controls.
· The one column form for controls is still included in the appendix, however, use of the two column form for controls is the preferred approach.
· The example used has been changed to reflect the two columns for controls.
· To determine the detection ranking, only detection controls are considered.
· For recommended actions, focus first on high severity and then on high RPN.
· Intent of actions is to first reduce the severity ranking then the occurrence ranking then the detection ranking.
Process FMEA – Introduction

Process FMEA – Development of a Process FMEA
· For Key Date, a note has been added for suppliers that recognize a due date at scheduled start of production is too late. The latest the due date could be for a supplier is the customer required PPAP submission date.
· In item 9, Process Function/Requirements, it is recommended to record the associated process/operation number for the step being analyzed.
· Additional examples of typical failure modes have been provided.
· In item 10, Potential Failure Mode, a note has been added that provides guidance on recording potential failure modes.
· For the severity ranking, the ranking that is used is the most serious for a given failure mode. If a failure mode has numerous effects, only the most serious matters.
· Severity table has been changed to recognize effects for both internal and external customers.
· A note has been added to the severity section.
· Additional examples for causes of failure have been provided.
· Occurrence table changed to failures per thousand vehicles/items.
· Appendix (Appendix I) provided with associated Ppk values to assist team in determining the occurrence ranking.
· Only two types of process controls are recognized:§ Prevention§ Detection
· For process controls, the form has been changed to provide for separate columns to record controls. One column is to be used to record prevention controls, the other column is for detection controls.
· The one column form for controls is still included in the appendix, however, use of the two-column form for controls is the preferred approach.
· The example used has been changed to reflect the two columns for controls.
· To determine the detection ranking, only the detection controls are considered.
· Detection table provides additional detail to assist the team in determining the detection ranking.
· For recommended actions, focus first on high severity and then on high RPN.
· Intent of actions is to first reduce the severity ranking then the occurrence ranking then the detection ranking.

Process FMEA – Follow Up
· Methods to use to ensure that concerns have been identified and actions implemented have been included.
Appendices
· Design FMEA Quality Objectives and Process FMEA Quality Objectives have been added as appendix A and B respectively.
· Both one column and two columns forms for controls are included.
· An appendix (Appendix F) has been added that provides information concerning the interfaces and interactions among systems, subsystems, and components.
· An appendix (Appendix I) has been added that includes a Ppk column for the Process FMEA occurrence table. It also includes sample calculations.

Best Regards,

Pete

[SniperMan]

A new edition of FMEA has recently gone out, we're a tier 2 supplier, is this new edition going to affect us. I read a rejoinder at the AIAG website that this new addition are only applicable to tier 1 suppliers of the big three. corporations..

Any comments

[Marc]

Hi, Pete! Well, I missed saying thanks for your post and attachment. It was posted during a pretty stressful time - right at that first week after the server software and host changeovers which was one of the worst projects I have ever had the displeasure to manage. What a nightmare...

Belatedly, thanks! Always good to see you stop by!

[Jim Biz]

I've just gone through (with our customers input) a very very detailed PFMEA exercise.

Question - If a probable failure is determined "at any given point in a process" and a similar action takes place in -(for example) four separate steps of the complete process - is it necessary to Duplicate the rankings for each operation?

We were struggling with - a "handling issue" and trying to determine what caused - scratches - nicks - burrs on one part at each step of a process --- what we ended up doing was duplication of all the Probable causes for each operation and then "re-ranking (or not) the Sev - Occ- etc.

Seems to me what we ended up with was nothing less than 650 line items if information - that consisted of 35-40 useable lines - and then duplicated to the point of sheer madness.

Is there a "rule/guyide for use" of some kind for an fmea that says - one probable cause - one probable effect - is enough?