Process FMEA - Need help to clarify O (Occurrence) and D (Detection) rankings

Jim Wynne

Staff member
Admin
#21
A nice post, Rich, and I especially agree with your last sentence. I'm going to disagree, however, with your advice regarding Detection ratings. In general, preventive controls are not intended to detect anything. To put it another way, nonconforming material can't be detected by prevention controls. In order to avoid confusion it's best to keep the concepts of prevention and detection separate. For this reason, the Detection rating should address the likelihood that a defective part will be detected either prior to being shipped or prior to having additional value added, given the present controls.
 
Elsmar Forum Sponsor
A

astudent

#22
I agree with both of you when you say that the initial qsmso's question is an important one, that's why I really wanted to understand the answer.
It has not been easy maybe for language problems as I'm not a native speaker of English (by the way , JSW05 when in message #5 you write "either the mode or the cause" do you mean both of them or just one out of the two?).
Now I think that I have understood how things go, thanks to RAHarpster's answer too.

As to the extimate of (O) from available data, I want to express my very personal thinking.
I write "very personal" because I've never heard people saying this, even if I thing it's an important topic to discuss about.
In statistics there is a golden rule: all the times you need to do an extimate, you need to plan (i.e. to design it) a poll and you cannot browse informations gotten for other reasons and make a late supposition!
A real example I know:
A company had sampling plans to test incoming products.
They followed a sampling scheme very similar to MLT STD 105E, so the sampling number was not a constant.
Every month they happend to make an extimate of their supplier rate of defects, and guess how did they extimate that rate?
From the the only information they had of course!
So if lot#1 had 10000 items and they checked 100 items out of them and just one was defective they assumed a rate of 1%.
Suppose lot#2 was in free pass, so it was not taken into account (what if the rate of defectives was 10%?); lot#3 had a sampling number of 10 and no one defective (rate 0%), and so on...
Just imagine what was the value of those statistics!!!
And a lot of high managers looked at those diagrams pretending to understand everything and making future plans based on them.
So be sincere... do you really think that most of the data, people referr to when they try to extimate (O) is better that data in my example?
Making a reliable poll is a very difficult science and you have to prevent all the possible causes of biasness.
I think that virtually all the infomations companies have are biased.
Maybe the reasons of biasness are very different from the ones I've shown but are always in place.
BTW, I don't think they could not be useful, you just have to know their limits, how to use them and expecially what the aim of FMEA is: that is not to make a scientific extimation of (O), but just to have an idea in order to prioritize actions.
Subjectiveness is fundamental, and data you have are just useful to help you having a more precise idea about the numbers.
The more you try to get objective your extimate from biased data the more you are wasting your time (and risking to abtain a less real extimate).
Just my 0,02.
 

Jim Wynne

Staff member
Admin
#23
astudent said:
It has not been easy maybe for language problems as I'm not a native speaker of English (by the way , JSW05 when in message #5 you write "either the mode or the cause" do you mean both of them or just one out of the two?).
You do pretty well in English. Is this what you're asking about?:
What you want to ask is, "How likely is it that this particular failure (either the mode or the cause) will occur?" Remember--it's possible to predict failure without necessarily knowing all of the possible causes.
If you are separately considering modes and causes, each should have its own Occurrence rating.

astudent said:
BTW, I don't think they could not be useful, you just have to know their limits, how to use them and expecially what the aim of FMEA is: that is not to make a scientific extimation of (O), but just to have an idea in order to prioritize actions.
Subjectiveness is fundamental, and data you have are just useful to help you having a more precise idea about the numbers.
The more you try to get objective your extimate from biased data the more you are wasting your time (and risking to abtain a less real extimate).
Just my 0,02.
I agree completely.
 
R

RAHarpster

#24
Prevention Controls and Detection Ratings

This is in response to the following post:
I'm going to disagree, however, with your advice regarding Detection ratings. In general, preventive controls are not intended to detect anything. To put it another way, nonconforming material can't be detected by prevention controls. In order to avoid confusion it's best to keep the concepts of prevention and detection separate. For this reason, the Detection rating should address the likelihood that a defective part will be detected either prior to being shipped or prior to having additional value added, given the present controls.

First, let me thank you for your comments. This is a very important topic and the more we can get people to think and talk about it the better.

I agree with the statement that non-conforming material cannot be detected by prevention controls. As you know, prevention controls when defined properly only look at the process. To a prevention control, the product is invisible. You also know that detection controls when defined properly only look at the product. To a detection control, the process is invisible.

The dilemma that we must deal with is when one reaches the goal we all want to reach and creates a process with perfect prevention controls that negate the need for any type of detection. Do we have a Process FMEA methodology that can properly assess the strength of a process in this state. The current Process FMEA detection rating system prescribed by most of the books in publication cannot handle this condition. Let me try to use a simple example to explain why this is so.

Process Action: Stamping Press
Failure Mode: Improper hole location.
Failure Effects: Cannot assemble part.
Severity: 8
Failure Cause: Operator improperly positions part in press fixture.
Occurrence: 1
Prevention Control: Auto sense of part position in fixture with auto shutdown. (This is a prevention control because it looks at the process).
Detection Control: None. (The prevention control will not allow the process to run when the process condition of improper part position in the press is detected. The prevention control is perfect thus negating the need for any detection control).
Detection Rating: If you follow the instructions in the currently published FMEA manuals you would have to assign a detection rating of a 10 since no inspection of the product for incorrect hole location is planned. Given the occurrence of 1 and severity of 8 this would result in an RPN of 80.

However, if an additional rating table is constructed that is based on the adequacy of the prevention control in preventing the product from being produced in the presence of the cause, one would have to assign the prevention control a detection rating of 1 since the prevention control is perfect. This will result in an RPN of 8 thus declaring that no additional effort is required in this area with respect to preventing improper hole location. To me, it is much more logical to assign an 8 to this line of the Process FMEA rather than an 80.

The majority of the books that are published state that the detection rating for a prevention control is embedded in the occurrence rating. This would be true if in all cases a process failure cause that is linked to multiple defects (failure modes) will cause all the defects at the same probability rate. However, we know that this is cannot be true since some defects are more sensitive to a particular adverse process condition than others. Because of this fact, the FMEA must provide us with a way of documenting the effectiveness of prevention controls. The logical place is the detection column but a separate rating system based on the probability of the cause happening that the control is trying to prevent is required.

Rich Harpster
Harpco Systems, Inc.
 
Last edited by a moderator:

Jim Wynne

Staff member
Admin
#25
RAHarpster said:
The dilemma that we must deal with is when one reaches the goal we all want to reach and creates a process with perfect prevention controls that negate the need for any type of detection, have we created a Process FMEA methodology that can support the condition.
A better question might be, having reached that lofty goal, where defects just don't happen, why would we need PFMEA at all?
RAHarpster said:
The current Process FMEA detection rating system prescribed by most of the books in publication cannot handle this condition. Let me try to use a simple example to explain why this is so.
While there's much about conventional PFMEA strategy that I disagree with, I think the reason for this is that the current ideas about PFMEA all assume that defects are possible. What's wrong with that?

Process Action: Stamping Press
Failure Mode: Improper hole location.
Failure Effects: Cannot assemble part.
Severity: 8
Failure Cause: Operator improperly positions part in press fixture.
Occurrence: 1
Prevention Control: Auto sense of part position in fixture with auto shutdown.
I think this might be where your theory breaks down; the assumption is being made that the prevention control in question is already in place. Had this been an actual PFMEA process, we would likely start with the mode and effects as you describe them above (I'd do it differently, but let it pass for now), then identify the potential cause(s) and a detection value for each. Given this information (in addition to the severity rating, of course) we could then consider the prevention control in question, and decide whether or not the probable risk justifies the expenditure. In this case, the prevention control might become a Recommended Action and if implemented the RPN could be recalculated accordingly.

Note that I'm not arguing with your basic point, which is that Detection factors are superfluous in light of defect-proof processes (you don't need to be able to detect what can't happen), but in such a case I might opt for just leaving the RPN alone altogether, because the same logic applies to the Severity and Occurrence factors--give them all a rating of 0 and move on. (Of course I understand that the current literature doesn't account for such a strategy, but who cares?)
 
R

RAHarpster

#26
The Process FMEA is a point in time audit of the adequacy of the process in producing within specification product. The occurrence number is the grade you are giving to the process when the controls listed in the prevention and detection columns are in place. Consequently when you place a control in a Process FMEA you are saying it exists. If the control does not exist, it should not be listed in the Process FMEA.

If a prevention control does not exist and the occurrence is high and you believe you have a prevention control that could help, you should first use the recommended actions column to define the new prevention control to be tried and evaluated. You should then evaluate the control and if the evaluation of the control proves that you can reduce the occurence rating, the control should then be implemented, placed in the prevention control column and the detection and occurrence rating columns of the Process FMEA modified to reflect the improved process. If your evaluation of the prevention control proves that the control does not do what you expected it to, you leave the recommended action on the Process FMEA so that no one else tries what you have already proved does not work.

Until an item listed in the recommended action column is evaluated, implemented and placed in the body of the Process FMEA, it remains only a "recommendation" and has no impact on the ratings of the Process FMEA. In 20+ years of reviewing Process FMEAs I have seen many people make the mistake of using the "Recommended Actions" column to store active prevention and detection controls. I have actually seen this in the last week when reviewing a client's existing Process FMEAs.

Rich Harpster
Harpco Systems, Inc.
 
Last edited by a moderator:
R

RAHarpster

#27
Two excellent points have been made during this discussion. The first was:
A better question might be, having reached that lofty goal, where defects just don't happen, why would we need PFMEA at all? The second was "What is wrong with assuming defects exist?"

In response to the first comment, you need the Process FMEA methodology to audit your process to determine if you have achieved the lofty goal of a defect free process. In response to the second comment, there is nothing wrong with assuming that defects exist. However, your task in using the Process FMEA process is to define the defects that are the most costly to the company and occur at the highest rates so you can define their root causes and properly target your resources to make the causes go away.

When done correctly, the savings can be quite large. When we first properly implemented the Process FMEA methodology in a plant that I ran 14 years ago, we were saving on average $144,000/week after nine months of implementation. The key to saving money using Process FMEAs is detailed definition of process failure causes and the implementation of prevention controls. Unfortunately, many Process FMEAs list the failure causes as "Operator Error", "Incorrect Setup" or "Equipment Malfunction".

Rich Harpster
Harpco Systems, Inc.
 
Last edited by a moderator:
Thread starter Similar threads Forum Replies Date
A Will Inspection Process need an FMEA? FMEA and Control Plans 7
I Need Help with Process FMEA Basics - Feasibility Analysis Outputs are Inputs to FMEA? FMEA and Control Plans 25
J Process FMEA Template with examples - Cold and Hot Forged components FMEA and Control Plans 4
C FMEA Process assessment In the Draft for the AIAG/VDA FMEA Manual is gone FMEA and Control Plans 0
Jimmy123 Example of a P-Diagram for Process FMEA - Uncontrollable noises FMEA and Control Plans 39
eule del ayre Merging FMEA and Quality Control Process Flow FMEA and Control Plans 0
K AIAG/VDA FMEA & Process Control Plans FMEA and Control Plans 0
S New AIAG FMEA Process - How to complete the new format FMEA and Control Plans 32
K PFMEA (Process FMEA) - Can be common for 3000 products? FMEA and Control Plans 2
D Where does FMEA fit in your ISO 14971 Risk Management process? ISO 14971 - Medical Device Risk Management 13
N In-Process Production Test Stations in PFMEA (Process FMEA) FMEA and Control Plans 18
N Process FMEA Occurrence Ranking for New Process FMEA and Control Plans 3
G AIAG FMEA - Major customers of the FMEA process FMEA and Control Plans 3
E How do I create a FMEA for the Internal Audit process? Internal Auditing 11
D How to fill a Process FMEA - Three columns - Severity, occurrence and Detection FMEA and Control Plans 16
L Injection & Extrusion Process FMEA FMEA and Control Plans 1
S Pilot Process FMEA Occurence Rankings FMEA and Control Plans 4
K Process FMEA Ranking Table for for Landing Door of an Elevator FMEA and Control Plans 7
G Severity in Process FMEA for Rework Process FMEA and Control Plans 1
R Is PFMEA (Process FMEA) for OEM sufficient to address Risk? ISO 13485:2016 - Medical Device Quality Management Systems 7
H FMEA sample for complete CNC machining process FMEA and Control Plans 1
T Software for linking Process Flow Diagram, Process FMEA and Control Plan APQP and PPAP 9
V Process FMEA for Stamping Process FMEA and Control Plans 5
G FMEA for Steel Tube Process for Automotive FMEA and Control Plans 4
L Most appropriate Auditor to check the FMEA & PCP of a Manufacturing Process IATF 16949 - Automotive Quality Systems Standard 5
C List Error Proofing in the Process FMEA or the Control Plan FMEA and Control Plans 2
J Process FMEA w/ Contracted Manufacturing ISO 14971 - Medical Device Risk Management 2
B FMEA example for Cokemaking Process at a integrated Iron and Steelmaking Plant IATF 16949 - Automotive Quality Systems Standard 3
S FMEA Process Step or Current Control FMEA and Control Plans 3
Z Pass-Though Characteristics in PFMEA (Process FMEA) FMEA and Control Plans 2
Q Chrysler required FMEA and Control Plan Document Audit Forms and Process Audit Check FMEA and Control Plans 7
V Examples of Design and Process FMEA for Pharmaceutical Product Development ISO 14971 - Medical Device Risk Management 7
S Galvanize and Chrome Plating Process FMEA wanted FMEA and Control Plans 2
S Product FMEA vs. Process FMEA - Aerospace Industry Machine Shop FMEA and Control Plans 5
R PFMEA (process FMEA) for Incoming Material stage FMEA and Control Plans 7
P PFMEA (Process FMEA) Recommended Actions help FMEA and Control Plans 9
J What is Incorporation in Assembly/Manufacturing Process in FMEA? FMEA and Control Plans 4
F Process FMEA for Welding and Laser Cutting FMEA and Control Plans 3
G Very Basic PFMEA (Process FMEA) Guidelines FMEA and Control Plans 17
R Process to use Special Characteristics as FMEA Inputs IATF 16949 - Automotive Quality Systems Standard 3
M PFMEA (Process FMEA) - What is the best approach? FMEA and Control Plans 14
F FMEA Optimization for our Die Setup Process FMEA and Control Plans 2
M PFMEA (Process FMEA) for Attribute Data Inspection Process FMEA and Control Plans 2
A Current Control Periodic Document Reviews - Process Updates in the FMEA FMEA and Control Plans 3
T PFMEA (Process FMEA) Recommended Actions Column FMEA and Control Plans 10
B Process FMEA (PFMEA) Same Multiple Causes for Different Effects FMEA and Control Plans 6
M Changing PFMEA (Process FMEA) RPN Values based on a Rejection FMEA and Control Plans 5
Chennaiite FMEA for Business Process Imported Legacy Blogs 6
K Does FMEA include Inspection Process? FMEA and Control Plans 11
Q Using a Custom Process FMEA (PFMEA) Form FMEA and Control Plans 4

Similar threads

Top Bottom