How do I define opputunities?

Morning(?) Helen,
It depends if you are looking at the macro or micro level at your process. At the macro level, with 8 processes, I would view the Opportunities for Defects to be 8 – as the product either comes out ‘good’ or ‘bad’ after each process.  At the micro level, looking at each of the 17 components, and its position/insertion/function/whatever there are another 17 opportunities – but within each of those opportunities, there may be further detail needed to define the problem more accurately. 
In other words, keep peeling away the layers of the onion until either your eyes water uncontrollably, or you get the data you need to fix the problem (the Y=f(x) bit, but each f(x) has its own f(x)s as well)
Batch quantity and the number of incorrect parts should only give you an indication of the likelihood of success 3/100=0.03 so success will be (1-0.03) = 0.97 or 97%.  With defects per unit now established as 0.03, multiplying that by 1,000,000 gives a value of 30,000
DPMO (Defects per million opportunities) = 30,000/No. of opportunities for error, which is now a process complexity biased value.
Does this help?
regards
James

Helen and Jaran,
Better still, click on the “Sigma Calculator” on this page, and below it you will see a link to “Learn how to define units and opportunities”.
This will also provide more detail.
regards
James

Helen,Here is my take:I make apart. My inspection sheet, developed based on the FMEA I conducted where I identified the potential failures for the product, has 10 characteristics. Every time I inspect the product, I inspect it for these 10 charateristics. Let’s say that 7 of the charatcteristics are continuous (can be measured) and 3 are discrete (a go/no go decision must be made). Further, let’s say that I produced 10,000 parts per work order. That would mean that there are 100,000 opportunities for error for each work order.
This is a pragmetic cut-off point. One could argue that all the characteristics on the drawing are opportunities for error.Marc Richardson
Sr. Quality Assurance Engineer

I don’t think you’ve missed a thing.
The trick is to identify all the things that could possibly go wrong with a unit. There are many ways of doing this and in my experience, none of them is 100% effective. Ussually, people agree on some practical standard of what is most likely to go wrong.
It is a commonly accepted practice to multiply the number of things that can go wrong with a single unit or transaction by the number of units or transactions being produced or executed to arrive at the number of opportunities. This can and does create large numbers of ‘opportunities’ against which a small number of defects can look impressive.
This is the basis of the charge that 6s practioners inflate the number of opportunities in order to show a 3.4 DPMO. Personaly, I prefer to stick with the familiar PPM measure. Low DPMO values are meaningless to customers who continue to receive product that does not meet or exceed their expectations.Marc Richardson
Sr. Quality Assurance Eng.

Marc,I would suggest this last post reflects your misunderstanding of opportunities.

Stan,I still don’t understand how you are defining opportunities. Please provide us with your formula for calculating DPMO and I’m sure that will clear things up. I’m always willing to learn.Thanks,
Marc Richardson
Sr. Quality Assurance Engineer

Hi Helen,
Sigma rating  =     No of defects  x  1,000,000
                        No of Units x  No of opportunites
For instance, if we consider the quality of an item of clothing sold at retailers, and out of 650 items sold, 2 fail for a broken zip and one for its’ general condition, then.
DPMO =    (2+1) x 1,000,000     =   1538
                          3 x 650
One other point… some texts would quote 769 as the answer here and not 1538.  This is because some scenarios exist where scores below the LCL are OK e.g. time taken for a letter to reach its’ destination – if it is less than the average, then this is good news!
Hope this helps.  Email me at [email protected] if I can help any further – I am not an expert, but I will try.
Regards
 
Dewi
 
 

Patrick,Point taken — see my post to Marc.

Opportunities are what you decide they are. If you are in a manufacturing world they are the dimensions on a part if you are in a service industry they may be customer KQC’s. But whatever industry you are in you must decide what they are and baseline it. Remeber the metric drives the action so if you have many many opportunites you will probably inflate your sigma values.
At Allied Signal it was decided that opportunities by definition was three times the bill of material. This precluded gamesmenship when it came to calculating sigma values.
In your scenario you would examine what operations were performed at each step of the process and what specification were affected. You would then calculate a first pass yield at each step, calculate your RTY and select your sigma value based in yield.
 

Hi Helen,
In your case you need to compute the output at each of 8 steps & get the sigmas at each step. Output of step 1 will be Input of step2.
Also according to me there are is only 1 opportunity for a qty to work. Either it is defective or it is not.
So as per me there can be only 1 opportunity for defect.
Most welcome for discussions,
Sophia

The measured part of the definition is nonsense.
One of the real financial benefits of achieving a high capability (choose your own definition here, but mine is Cp>1.67, Cpk>1.5, and Cpm>1.33) is the tremendous reduction or elimination of the non value added activities called test and inspection.
Using the measured definition, opportunities would decrease as capability increases which makes no sense at all.
This definition comes from the active vs passive arguement put forth in The Vision of Six Sigma by Harry oublished in 1994 (lovingly referred to as the white catoon book). Just know that the person making this arguement thinks that the DMAIC method is superior to Lean for reducing time and inventory.

I hope your masters work is more rigorous than your ability to calculate opportunitities.
From the information given opportunities are at most (8+17)*100

Hi,
DPMO equals (number of defect divided by number of units)x1000000 and this divided by the opportunities per unit.
So in your case:
DPMO = ((3/100)x1000000)/ 17 = 1764