The company I'm currently with begins their Control Plan prior to beginning their FMEA. After they identify critical characteristics on the CP, they do FMEAs on the critical characteristics. We have a new person at our company that says we CAN NOT start our Control Plans until our FMEAs are complete.

Why Not? Am I missing something?

Appreciate your comments.

The sequence of events is: Design FMEA, Process FMEA, and then Control Plan. FMEAs determine what characteristics need to be controlled and the Control Plan describes how to control it. Ford's Dynamic Control Plan actually combines the PFMEA and the Control Plan.

I agree with Spaceman. I was really shocked when I first read the starter. I have been doing my FMEAs for the last couple of months and was listing down the critical characteristics so as to be included in the control plan and what I read was exactly the opposite of what I was doing.. I am not still sure about whether the sequence specified by Spaceman "DFMEA, PFMEA and control plan" is the right one, but as far as my knowledge goes, its FMEAs first and then the control plan. As described, come out with your process flow chart and diagrams, do FMEAs for all your processes and also your designs, identify the critical processes and then include these critical processes in your Control Plan to control them. Thats what the name suggests.
Any comments about this SENIORS???

Igenerally start with a control plan - enter customer critical characteristics. Then DFMEA, then the Trilogy (PFMEA, CP and Process Flow.

Have you reviewed http://Elsmar.com/pdf_files/APQP_Q.pdf ?

How do you identify the critical characteristics Marc? I already told the way I have planned to do it and have been doing it for the last two months. Am I going wrong? Is this not the best approach? Do you suggest me to change my approach and catch another path?

Your control plan evolves. As I said, I start a control plan first and put customer defined critical characteristics on it. To get your critical characteristics you look at everything from industry standards on. Y9ou might machine a part and for that part to go in an assembly it may require a hole be placed within a certain area. It may be that area is critical, or it may not. You simply have to use your judgement and experience. There is no list you can go to.

I don't think the path is that critical. There is almost always overlap - as the DFMEA is being done quite often a PFMEA team (or person {reality, folks} is doing the Process FMEA. You have to understand the process, however. Sometimes there are DFMEA entries that are carried over to the PFMEA (not often, but it does happen from time to time).

I prefer to have a projected process flow diagram before starting the PFMEA as I believe you have a better idea of what to look at.

I wouldn't let the sequence bog you down. If you do the PFMEA before the DFMEA (which is very, very common - but I would say it is not the 'smartest way' to do it), it's a matter of reviewing the two together and checking to see if anything from the DFMEA should be added to the PFMEA.

If you review the APQP document I linked above and still have questions, come on back and ask more questions.

Gotta run. I'm leaving for Tampa in a couple of hours.

We always did PFD, PFMEA, then PCP. DFMEA was provided by design as the first document. In reality the PFMEA requires you identify "current process controls" so it was in all practicality a simultaneous activity. Alot depends on whether you are looking at an existing process or developing a new one.

Marc, it's not often I disagree with you... the customer's critical characteristics should be the "Input" in Phase 1 of APQP. Those characteristics are part of the design consideration, thus included in the Design FMEA.

The APQP manual provides guidance on the sequence of events for document preparation;
DFMEA,(PFD,PFMEA) and Control Plans.

My 2 cents:

While performing an FMEA, you need to have some idea of how a failure mode is controlled, even if the FMEA comes first in your sequence of events. Additionally, is a low RPN value as a result of the preliminary (?) control plan activity?

Regards,

Kevin

(May be I don't understand the discussion quite well.) I think an efficient control plan should focus in the first place on "critical" issues. These critical issues can be identified by an FMEA. The most logical order would then be FMEA, next control plan. Bye the way by nature of an FMEA it results in an action list, what could well be a control plan. Without a critical analysis your control plan hangs in the air, I'm afraid...

Kevin, RPN reduction is normally through design and/or process improvement at the pre-production phase. However, since FMEA is a living document, you are absolutely right in saying FMEA RPN reduction is result of control plan.

Flowcharts come first; FMEAs are created from the flowchart; control plans are created from the FMEA.

Originally posted by Dawn:
Flowcharts come first;Sometimes at the Design FMEA stage the process flow is not known or defined. As Laura said: "...it was in all practicality a simultaneous activity..."

The chicken came first...

Originally posted by Spaceman Spiff:
Marc, it's not often I disagree with you... the customer's critical characteristics should be the "Input" in Phase 1 of APQP. Those characteristics are part of the design consideration, thus included in the Design FMEA.I'm sorry if I didn't make that clear. Before you do anything you have to get your customer requirements - which is why a control plan (sequentially) is an early (sometimes the first) document. And remember, the control plan, like the other documents, 'evolve'. In QS there is a requirement for 3 'levels' of the control plan (sometimes all three ar the same but that's another topic...). Whether or not you start the control plan right away, or after you do the design FEMA, you need to know the customer requirements including critical characteristics. You can't (in my opinion) do a design FMEA unless you have the customer requirements including critical (and safety, etc.) characteristics.

Folks,
I'd say find out customer requirements, use a QFD or cause and effect matrix to decide what process or design inputs have most effect on these requirements, then do your fmea(s) to get to potential failures.
After all this good stuff you'll know what to include on your control plan.
Where the confusion sets in is that the control plan will quite often include some product based controls that were specified by the customer right from the start.
If you want to pencil these in on a control plan form, right from day one -what the heck!
Who cares as long as you end up with control of the product characteristics that the customer specified, plus all the process based inputs that will have an effect.

Cheers

Brian

Originally posted by Spaceman Spiff:
Marc, it's not often I disagree with you... the customer's critical characteristics should be the "Input" in Phase 1 of APQP. Those characteristics are part of the design consideration, thus included in the Design FMEA.Agreed.

Folks,

All books and literatures about Control plans would tell you that it is generated after your FMEA. Your FMEA should drive your control plan since your critical and significant characteristics shall be defined during FMEA brainstorming. The best way is to make a SIPOC , then FMEA , then the control plan, then do a proactive review of your leading indicators, then review and update the FMEA again and again..

How do you identify the critical characteristics Marc? I already told the way I have planned to do it and have been doing it for the last two months. Am I going wrong? Is this not the best approach? Do you suggest me to change my approach and catch another path?

u man u want to become defective when u r already right.DFMEA covers design risks that were earlier covered by design review only.

1. very 1st step is identify process- Process flow
2. then evaluate the process for failure moded - PFMEA
3. generate Critical / special characteristics
4. Identify controls for failure modes & special for special characteristics
5. Prepare quality check points

now change the flow & tell me what you will get if u by-pass anyone

I'm sorry if I didn't make that clear. Before you do anything you have to get your customer requirements - which is why a control plan (sequentially) is an early (sometimes the first) document. And remember, the control plan, like the other documents, 'evolve'. In QS there is a requirement for 3 'levels' of the control plan (sometimes all three ar the same but that's another topic...). Whether or not you start the control plan right away, or after you do the design FEMA, you need to know the customer requirements including critical characteristics. You can't (in my opinion) do a design FMEA unless you have the customer requirements including critical (and safety, etc.) characteristics.

Marc:

I agree with you completely on this one. Control Plans under TS still have the three (3) levels.

I could never figure out how anyone could go into a Process FMEA (real one and not the fakes 1 day be sample submission) without having a draft Control Plan. I agree that the final Control Plan will be the outcome of the Process FMEA but not having anything on paper going into a meeting just doesn't make sense to me.

Folks,

All books and literatures about Control plans would tell you that it is generated after your FMEA. Your FMEA should drive your control plan since your critical and significant characteristics shall be defined during FMEA brainstorming. The best way is to make a SIPOC , then FMEA , then the control plan, then do a proactive review of your leading indicators, then review and update the FMEA again and again..

I agree this is the generally correct approach, but do bear in mind they both kind of develop concurrently, depending on what information is available and when. But, yes, in finalizing, the PFMEA should be finalized first, because it directs the control plan.

:topic:

Reading this reminds me of the chicken and egg story ;)

The sequence of events is: Design FMEA, Process FMEA, and then Control Plan. FMEAs determine what characteristics need to be controlled and the Control Plan describes how to control it. Ford's Dynamic Control Plan actually combines the PFMEA and the Control Plan.

I'd look at current D,P FMEAs and control plans from a similar product first, saves alot of time... Without DFMEA you cant really set specs for parts, without clearly specified part process cant be defined, part and process not defined nothing to control!

I agree this is the generally correct approach, but do bear in mind they both kind of develop concurrently, depending on what information is available and when. But, yes, in finalizing, the PFMEA should be finalized first, because it directs the control plan.Wow! reviving a seven-year-old thread!

FWIW (for what it's worth):
I always thought of the FIRST step as Contract Review where supplier and customer reach agreement on the exact requirements for the product or service.
For the benefit of newbies and latecomers:
DFMEA = Design Failure Mode and Effects Analysis
PFMEA = potential (OR process) failure mode and effects analysis
FMEA = Failure Modes & Effects Analysis
PPAP = Production Parts Approval Process OR Program
APQP = Advanced Product Quality Planning
Control Plan = the overall plan for producing a conforming part or service, which may include minutiae such as speeds, feeds, machine and jig design, in-process inspection, final inspection, packaging, labeling, documentation, etc. etc.

(ignoring specific industry requirements for PFMEA, FMEA, and PPAP), it seems to me the next logical step after reaching agreement on the requirements for the finished product or service, including critical characteristics of form, fit, and function, the supplier maps out a preliminary plan for producing the product or service. As part of that plan, the supplier asks, "Where can this plan screw up and produce non-conforming service, material, or scrap (PFMEA and FMEA)?" After making decisions about the frequency and severity of those possible screwups, the supplier concocts a final control plan, considering the most efficient and profitable method of producing the conforming product or service after taking into account all the soft costs of dealing with any nonconforming parts that may slip through (returns, rework, scrap, recall, lawsuits, government sanctions, customer loss, etc.)

Sometimes, despite the best planning, the process still has glitches, so folks have to go back over every step. The real answer, it seems to me, is that the entire process of Contract Review, FMEA, PFME, FMEA, APQP, PPAP, Control Plan is continually fluid, always subject to improvement or "tweaking," continually searching for maximum efficiency and thus maximum profitability. Therefore, which exact stage of the process where one begins is moot because it's all part of one big circle, just like Deming's PDCA (Plan, Do, Check, Act.)

Your control plan evolves. As I said, I start a control plan first and put customer defined critical characteristics on it. To get your critical characteristics you look at everything from industry standards on. Y9ou might machine a part and for that part to go in an assembly it may require a hole be placed within a certain area. It may be that area is critical, or it may not. You simply have to use your judgement and experience. There is no list you can go to.

I don't think the path is that critical. There is almost always overlap - as the DFMEA is being done quite often a PFMEA team (or person {reality, folks} is doing the Process FMEA. You have to understand the process, however. Sometimes there are DFMEA entries that are carried over to the PFMEA (not often, but it does happen from time to time).

I prefer to have a projected process flow diagram before starting the PFMEA as I believe you have a better idea of what to look at.

I wouldn't let the sequence bog you down. If you do the PFMEA before the DFMEA (which is very, very common - but I would say it is not the 'smartest way' to do it), it's a matter of reviewing the two together and checking to see if anything from the DFMEA should be added to the PFMEA.

If you review the APQP document I linked above and still have questions, come on back and ask more questions.

Gotta run. I'm leaving for Tampa in a couple of hours.
Marc,
Agreed to your suggestion on the decision to do that way. It is fine as long as the technology is a carry over one, like maybe the transmission of the vehicle, anything from existing production. Say if there is a New concept like a Hybrid technology, then you want to follow the sequence of DFMEA, DVP & R, DFMA(virtual design), then the process documents (PFD,PFMEA and Control PLan).
We have seen both the scenarios like carryover technology and new technology.
Thanks
Mustafa Siddiqui