# Can You Have More Defects than Opportunities?

Six Sigma – iSixSigma Forums General Forums General Can You Have More Defects than Opportunities?

Viewing 17 posts - 1 through 17 (of 17 total)
• Author
Posts
• #53391

Dunderdale
Participant

I recently attended a six sigma training course. I got into a debate when the tutor gave the following example.

“When producing a car there are 4 opportunities for defects (Bodywork, interior, wheels and engine). If the car have 3 dints in the bodywork, 4 rips in the interior and 3 flat tires there wouldbe a total of 10 defects

In this situation the DPMO would be 2 500 000″

I argued that it is impossible to have more defects than opportunities and that the concept of defective boolean so in the situation you would have 3 defects and a DMPO of 750 000?

I went on the try and back up my argument but I think that this is enough information for my question to be answered.

Who is right?

0
#189929

Hody
Member

You are both correct in different ways… The tutor was calculating DMPO right, but I don’t think he (she) was correct to apply it in that scenario…

If you want to measure DPMO, then you have to check each opportunity for each sampled unit. But you only count each opportunity once…

The issue that you have raised is probably because that mfg line should track DPU instead of DPMO. For DPU, use a u-chart to track boundless defects within a unit (or time frame). DPU and u-chart are based on the Poisson distribution, and DPMO and p-chart are binomial-ish. For DPMO, use a p-chart so it will translate defect rates to 0-100% (not more than 100%).

I hope that helps…

0
#189931

Dunderdale
Participant

I think so.

So what I calculated was the DPMO and what the tutor calculated was the DPU (defects per unit?)?

0
#189932

Hody
Member

For a single car failing bodywork, interior, and tires but not engine is 3 defects for 4 opportunities. DPMO = 1,000,000 * 3/4 = 750,000

For a single car there were a total of 10 defects in that 1 unit. DPU = 10. Common examples of DPU are like blemishes per washer top, scratches per sheet of glass, or nicks per 1000 ft of wire,…

0
#189946

Szentannai
Member

Hi,
I think this is a very interesting question. My 5 cents would be that the error is that the way the opportunities and the defects are calculated is different.

I think that an opportunity is a binary variable – either it is defective or not. So, it should IMHO be impossible to have more defects then opportunities and I believe the trainer was wrong.

If you wanted more details then simply to say the wheels are defective (have one or more defects) you could define a finer division – for instance front left wheel, front right wheel etc. In this case you could have 4 opportunities for wheel defects and if you find (any number of) defects on two wheels but none on the other two then your DPMO would be 2*1000000/4 that is 500000.

If you wanted even more detail you could divide each wheel into quadrants. Then the number of opportunities will be 4*4 = 16 and you can count the defects accordingly.

Same logic could apply to body or any other part.

Regards
Sandor

0
#189953

Rebin Raju
Participant

Hi,

This is an interesting discussion thats applicable to daily implementation in processes across.
My opinion with what I have dealt in my Quality experience is,
1) There cannot be more defects than opportunities
2) DPMO definition itself says defects per million opportunities (DPMO equals the total number of defects per unit divided by the total number of opportunities for defects per unit, multiplied by 1,000,000)
3) So for any process or product to have or identify defect, first opportunity for defect should be defined. If actions performed for meting requirements of the defined opportunity is not met, then that opportunity turns out to be defect

In this example, I guess Tutor failed to define opportunity for error/defect and that lead to this discussion.

Regards,
Rebin

0
#189954

Cone
Participant

It is possible to have more defects than opportunities. Two scenarios –

1) Rework is ineffective, a repair occurs multiple times.
2) Opportunities are the value added things that must go right. A simple example is a piece of information on a form. Only one way to right, but it can be omitted, transposed, filled with a wrong code, …

0
#189955

Szentannai
Member

Hi,
I disagree.

Rework/repair should be considered a different opportunity IMO. It is definitely a different error scenario than the original one.

In case of a form it will depend on how you count defects. If any collection of defects that will lead to the incorrect filling of the information on the form is counted as one, then you have one opportunity and one defect. E.g. the wrong code is written, transposed.

If you count this as two defects then you imply that you have two separate opportunities. Again, the number of defects will be smaller or equal to the number of opportunities.

The point is, that with this definition DPMO is a multiple of the estimate of the probability of making a defect.

0
#189956

Fontanilla
Participant

I think the tutor is wrong. The value of defects and opportunities is not an intellectual exercise but needs to point to real-world solutions. The “art” of dpmo is not defect counting, that’s the easy part, it’s the opportunities.

Assuming the goal here is to focus limited resources on problems, the opportunities for “dints” in the bodywork should be a count of all the manufacturing operations where those defects could occur (common cause) and an understanding of where in the process the defects occurred. Yes, this is really hard, but it’s how it’s supposed to work. To count bodywork as a single opportunity for defects is ludicrous unless the body is purchased, painted and whole and is installed in a single operation.

The danger of mapping opportunities to the factory floor is the risk of establishing a ludicrously high number of opportunities so that 3 dings looks like “six-sigma quality levels”. Defects as perceived by the customer as negative differentiators of quality are unacceptable, regardless of the guality level. Don’t tell me I’m buying a six-sigma quality car with dings and tell me it’s my perception problem!

0
#189959

Jonathon
Participant

In control charts there is a U chart covering defects per unit, which readily accommodate more than one defect. From a standpoint of computing so-called process sigma, though, the opportunity space (area under the normal curve) is defined to be unity. The statistical wonks can debate at length about handling a defect count that exceeds the opportunity space. Personally, I’d be less worried about the way to report the numbers, and I’d want to focus on what’s causing the defects.

We need to be careful not to “inflate” the opportunity space. An erroneous approach is to count the types of potential defects as separate opportunities. Fortunately your tutor has not done that.

Also, rework and repair may be necessary, but they should not be included in opportunity counting. That’s because they are non-value-added steps that should be eliminated once the process is able to deliver right-first-time quality.

0
#189960

Anonymous
Guest

Dan,

I think your example of a car’s chassis is excellent. It is similar to a web or the area of a semiconducting water.

To my mind, the real cause of confusion is the difference between a point defect, a line defect, and an area defect, such as a dig on a lens surface, a scratch on a vehicle’s duco layer, or a scuff on an optical storage disk.

This is why many Motorolans still prefer to use Yield, % Defective, or Defects per unit area – not defects per unit.

Defects per unit is incorrect since to my mind because it does not satisfy the assumptions of the Poisson distributoin.

Regards,
Andy

0
#189968

Sajulal S L
Member

Hi,

I fully agree with Sensei…Just to add my bit here…

The tutor is confusing between ‘defect’ & ‘defective’.
Defect – Any failure to meet the specification / customer requirement
Defective – An item / unit with one or more defects.

There can be multiple defects on the same defective part

0
#189971

Mike Carnell
Participant

Of course you can.

The opportunity count to do something os the opportunity count to do it correctly. This is a basic concept that most seem to miss. It is not the number of ways you can do it incorretly.

When you make a solder joint there is one opportunity to do it correctly. The is the possibility of having insufficent, excess, pinholes, pits, voids, nonwetting, dewetting, etc. If I have a pinhole and insufficent solder then I have 2 defects and only 1 opportunity.

When you count the number of possible ways you can screw it up you are inflating the denominator and turning you data into pure nonsense. Continuous improvement is a numerator job not a denominator game. It also is not some philosophical exercise – opportunity counting is basic. If you struggle with this concept and waste your time in a discussion about it you should be working in some other area.

Just my opinion.

0
#189973

Szentannai
Member

Carnell: “If you struggle with this concept and waste your time in a discussion about it you should be working in some other area.”

Or, at the very least you should go to a discussion forum where discussions are tolerated.

0
#189985

Mike Carnell
Participant

Sandor,

It isn’t about tolerance and pontification. It is about being so far out of the loop on such a basic concept you aren’t relevant.

Just my opinion

0
#189987

HBGB
Participant

Sandor, that was weak

0
#190048

Szentannai
Member

Well, I happen to disagree with Mike on this point, but IMO if the premise is that you should only be in this discussion IF you agree with him (on this particular point) then there is no point in discussing it at all.

However, I think the whole point of a discussion forum is to discuss things – so…

0
Viewing 17 posts - 1 through 17 (of 17 total)

You must be logged in to reply to this topic.