# DPMO Calculations

Six Sigma – iSixSigma Forums Old Forums General DPMO Calculations

Viewing 15 posts - 1 through 15 (of 15 total)
• Author
Posts
• #34820

SSNewby
Member

I’d appreciate it if I could get some guidance on the calculations for DPMOs and the when and why of their use.  As I see it for Defects Per Million Opportunities (use First Pass Acceptance only) DPMO is used to Normalize when comparing:  1.) different products in same process  2.)  different products different processes  or  3.) same products different process

For DPMO calculations:
2.) Look next to CTQ Critical to Quality activities intended to meet the CTC objectives
3.) Determine the points throughout the process that control the CTC > CTQ
4.) Calculate the number inspection and control point opportunities for failure
5.) Calculate the cumulative number of first time failures and passes (for all process steps, start  finish) and multiply both by the number of opportunities, divide cumulative failures by the cumulative number of units submitted, multiply that by 1 million.

This would say that a separate DPMO is derived for each CTC criteria using a aggregate of all of the CTQs that support the CTC and factoring in all of the Opportunities for Failure that relate to the CTQ.   An overall aggregate DPMO for the product or service could then be derived using all of the CTC DPMOs.  And this would/could be done regardless of existing QA schemes for AQL, RQL, etc.   Does this sound appropriate?

Thanks for helping,
SSNewby

0
#96437

Mikel
Member

Opportunities have nothing at all to do what inspections, tests, or controls you have or don’t have in place.
Opportunities are the count of the value added transformations your product or service goew through to make it right.
DPMO is a ratio (times a million) of things gone wrong to things which had to go right. Nothing more. Opportunity counting is a detailed process map of all transformations. Value added counts, everything else does not count. Tests and inspections are, by definition, not value adding (cost avoidance at best, but quite often superstitious dances).

0
#96438

SSNewby
Member

Thanks for responding.  I was hoping to get your input amongst a few others.
When you complete your process map and have identified your transformation points, your test/inspection points become observation points and tally opportunities.  Assuming (and it’s a broad assumption) that effective techniques have been used to identify appropriate transformation control points and are then used as test/inspection (failure observation) points for DPMO calculations, do you then see inconsistencies in the approach?    I live and work in a very risk averse environment stressing a high organizational internal locus of control and much time is spent in process review and establishing appropriate process observation and control points – so I believe that we have a pretty good comparative baseline for process control.   But, I’m also convinced that many opportunities for process improvement exist.

0
#96439

Gabriel
Participant

Stan,
As you probably noticed already, I am not used to opportunities count.
After your explanation, I would like you to clarify the following to help me understand:
– Does “components” count as opportunities?
– Is each manufacturing step (like machining, assembly, etc) one and only one opportunity?
– If the last one is true, that means that the same number of opportunities are count (allways one) for example for a simple (few CTQs involved)  machining operation (like drilling a hole where you only care about diameter) and a more complex one (like grinding the groove of a ball bearing where several CTQs are involved like groove diameter, groove radius, roughness, waiveness, paralelism, ovality, jut to mention a few ones). Not to say a multispindle turning operation or a transfer turning machine. I worked in processes where more than 50 characteristics (ok, not all of them CTQ) where defined in “one step”.
– In an assembly operation, does “Mounting each component, tightening each nut, etc” count as individual opportunities or is just one for the value added operation of “assembly”?
– Finally, how is a “defect” counted? The typical definition of “defect” is “the failure to meet a requirement”. With that definition and your definition of opportunity, you can have more defectes than opportunities! If a DPMO of 1 million equals a sigma level of -infinity, I cant’t imagine what a DPMO greater than 1 million would be (in fact, sigma level would be not defined). My (poor) understanding was that there is an opportunity where there can be one and only one defect. If not, I do not understand the nature of the ratio defects/opportunities. Shortly, can I do 2 things wrong after 1 opportunity to make it right? Out of 10 things that had to go right, can 20 things go wrong?
Thanks.

0
#96448

Mikel
Member

control points are for defects, they are not opportunities.
Think of it this way. The better your process the less control / tally points you need. By your method opportunities just went down, sigma level also went down – it does not make sense.

0
#96449

Mikel
Member

You can absolutely have more defects than opportunities.
Each purchased component is one opportuniy (same logic as each product to your customer is one opportunity).
Each and every transformation is an opportunity. Your example of tightening is a good one. You have four screws to tighten, each screw is an opportunity and each tightening is an opportunity – so a total of 8. If you go back to check torque, it is not an opportunity, you already had one of doing it the first time, that you are not sure is inspection and by definition non-value added.
You question of what is the sigma level where defects are greater to or greater than opportunities is only an interesting academic question. If you are that bad go improve something – anything. It would be hard to go wrong. Sigma level discussions are really only interesting when you start to get good – say better than 4.5 sigma.

0
#96462

SSNewby
Member

I hate to get caught up in circular logic, but following the example that you are using with four screws and four tightenings of the screws, giving eight opportunities for a CTC need of in place and properly torqued screws to exist, a means of knowing that you met your CTC/CTQ goal could be through someone having checked, post-tightening of the screws.   If your quality system has a person checking the CTC/CTQ criteria of screws in place and properly torqued, you have an observation /tally point to use in your calculation of DPMO.  You had an opportunity to meet your CTC/CTQ goal  what was the result?  I see that as separate and apart from the goal of improving your quality yield to the point of not having to check whether or not the screws have been installed and tightened.   The first is the creation of the metric and the second a logical derivative action/result of having effectively used the metric.

0
#96463

Gabriel
Participant

Just to make it more explicit, help me with this simple examples:
1) All components are bought, they are all different, and the following assembly/packaging operation are all manual and made by the same person. Parts A and B are tighten together with 2 screws C and D. A shaft E is inserted in a hole in the part A. This assembly is put in a plastic bag F together with another screw G that the customer will need for instalation. Finally, a pre-printed label H is stuck on the bag. The opportunity count is 8 components + 2 screws to tighten + 1 shaft to be inserted + 1 packaging (both the assembly and the aditional screw must be included) + 1 label (the right one) to glue = 13 opportunities. Is that right?
2) The same assembly and packaging is done in an automated machine. I feed it with a coil of plastic bags, a coil of stickers, and 6 drums, each one containing a lot of each of the other 6 components. The machine delivers the product assembled and packaged. If working correctly, the machine will stop when low of any component or packaging material. Opportunities = ?

0
#96468

Mikel
Member

I agree with 1. 2 would be no different – the things which were needed did not change just because the operation was automated.

0
#96469

Mikel
Member

I agree with you as long as the opportunity count doesn’t change from 8. In other words, I did not inflate the opportunities because I did not have a means of knowing they were right and had to do a check.

0
#96472

SSNewby
Member

I gotcha.  Thanks again for the help.

0
#96485

grover
Participant

Stan, I have a question regarding ” opportunity/ies” with respect to the service industry. What is the criteria of selecting appropriate, right and to-the-point opportunity/ies right the first time in  the DMAIC proces, since it directly effects DPMO and hence the Sigma value. For instance how do we exactly define opportunity/ies  in a service industry where customers are served by bunch of employees of the company.
Regards

0
#96495

Anonymous
Guest

Stan:
I recently came across in interesting case very simular to your simple example. A manufacturer of hospital entertainment systems mounted an LCD screen into a patient’s monitor using the wrong torque setting so that the ‘ground plane’ inside the LCD became damaged. This periodically caused white lines to flash across the screen and really upset the patients – even more than the ambulance chaser advertisements. This raises the spectre of having four opportunities, and four defects with one cause?
PS: I fully support your attitude to problem solving – who cares how it is done, as long as it gets improves, and the sooner the better!
Andy

0
#96525

grover
Participant

Stan, I have a concern with the answer you gave to Gabriel regarding the # of opportunities. Gabriel was talking about one simple part with eight opportunities in it. It was comparitively a simple case in order to realise # of opportunities. How about a very complex system. How can we say about a complex system that we took care of all possible opportunities in it. If the actual # of opportunities are far more / far less than what we determined, then it will really distort our DPMO and hence the sigma value!!
Also, i was concerned about the # of “opportunity/ies” with respect to the service industry. What is the criteria of selecting appropriate, right and to-the-point opportunity/ies right the first time in  the DMAIC proces, since it directly effects DPMO and hence the Sigma value. For instance how do we exactly define opportunity/ies  in a service industry where customers are served by bunch of employees of the company.
Regards

0
#96526

Mikel
Member

This is simple (sometimes tedious in the case of complex products or services)
One opportunity for every procured part, service, information, etc. as long as it goes to what is important to the end customer.
One opportunity everytime your organization transforms the product or service in a way that is important to the end customer.
How do you know it is transformed – it is different than before and closer to the thing your customer wants.
What is not transformation? Look to Lean for the definition of Muda – they are all non-value adding and therefore not opportunities.
If you want to do something different and count your checks, inspections, tests, audits, … as opportunities? Simple – you don’t get it. This method will get people to add these non-value adding steps – just the opposite of what we are trying to achieve.
What are we trying to achieve? It is not Six Sigma. It is competitiveness and responsiveness. Six Sigma is a set of tools and thought processes to get you there. If any part of the tool set does not drive competitiveness and responsiveness, it is a bad tool and does not have a purpose.
I am done with opportunities. Go figure out the system and make sure all of the pieces fit. Opportunities based on checks does not fit.
Done

0
Viewing 15 posts - 1 through 15 (of 15 total)

The forum ‘General’ is closed to new topics and replies.