# PPM Calculation. Making it useful

Six Sigma – iSixSigma Forums Old Forums General PPM Calculation. Making it useful

Viewing 28 posts - 1 through 28 (of 28 total)
• Author
Posts
• #51831

Renaud Bompuis
Participant

Hi,I’m having an issue with PPM calculation for defective parts received from supplier.
The current calculation is based on the monthly volume of parts delivered by a given supplier and the volume of defective parts (either found during incoming inspection or later during production).What bugs me is that supplier deliveries vary a lot month-to-month and defects may be found in the following months only.The issue of course is that if we only consider monthly volumes of received/defective parts we get PPMs that are all over the place, for instance:JAN2009:
– Defective: 0 parts
=> monthly PPM: 0
FEB2009
– Defective: 10 parts
=> monthly PPM: 10,000,000 !!
MAR2009
– Defective: 20 parts
=> monthly PPM: ?? (0?) monthly PPM: 0
FEB2009
– Defective: 10 parts
=> monthly PPM: 10,000,000 !!
MAR2009
– Defective: 20 parts
=> monthly PPM: ?? (0?) monthly PPM: 0
FEB2009
– Defective: 10 parts
=> monthly PPM: 10,000,000 !!
MAR2009
– Defective: 20 parts
=> monthly PPM: ?? (0?) monthly PPM: 10,000,000 !!
MAR2009
– Defective: 20 parts
=> monthly PPM: ?? (0?) monthly PPM: 10,000,000 !!
MAR2009
– Defective: 20 parts
=> monthly PPM: ?? (0?) monthly PPM: ?? (0?) monthly PPM: ?? (0?)To me these figures mean nothing: we can’t really act on them.Even using a sliding window and using the sum of received items and defective items over say the last 12 months would show large variations in the PPM.I thought that a more informative way of using PPM would be to use the overall historical amount of parts received/defective from this supplier and recalculate that value very month.
I thought also that showing a rate of variation (difference between monthly PPM) could be helpful in determining if a supplier gets better of worse.Practically, what would you propose in such cases?Thank you for any response.

0
#180938

Renaud Bompuis
Participant

For some reason my example data got mangled, trying to reformat it another way:JAN2009: Received: 1000 parts – Defective: 0 parts => monthly PPM: 0 monthly PPM: 0 monthly PPM: 0 FEB2009: Received: 1 parts – Defective: 10 parts => monthly PPM: 10,000,000 !! monthly PPM: 10,000,000 !! monthly PPM: 10,000,000 !! MAR2009: Received: 0 parts – Defective: 20 parts => monthly PPM: ?? (0?)

0
#180939

Caesar
Participant

What is so significant about a month? Why not just report %Defective for a constant sample size? You could also use a rolling sample size if it helps solve the problem. What is the purpose of your data collection?Are you sure your criteria is the same as the supplier?Do you both have the capable measurement system?Are all drawings to the correct revision?Bonne chance ..

0
#180942

Renaud Bompuis
Participant

Thanks for the reply Caesar.The point is that I’m an external party and the factory got used to calulating the PPM this way because the main corporate office wanted it that way.
The issue is that this monthly figure is not useful to them and I was trying to get a better indicator that could be useful for triggering some corrective action.I like the idea of he fixed sample size but the problem is that defects reported may not come from the sample, making our PPM go jump all over the place again without adding much meaning to the values we get.Most of our suppliers don’t even have PPMs, we’re talking about a factory in China that uses Chinese suppliers for the most part.

0
#180944

Caesar
Participant

They say the weather is nice in Beijing this time of year :-)

0
#180945

Renaud Bompuis
Participant

Too cold for me. I live in Hong Kong and prefer the weather here at this time of the year :-)

0
#180958

CT
Participant

Is there no way of matching the error to the order?
Also if there’s a lag can you have the error rate against the previous months figures. This is usually approx and depends on usage rate but I’ve seen it used to give an indication of issues.
As a minor point are you looking at stock reduction as you seem to carry a lot of stock which you don’t use for several months.

0
#180988

Jon Heller
Participant

From my experience PPM calculation should be made as a percentage of devices manufactured during the production cycle or lot. so if you receive 1000 pieces with zero failures, it is not necessarily a 0.
You need to ask how many parts were manufactured and what the failure rate was in production for a PPM. You may well have received a shipment that went through some QC or QA where failures are filtered out.
You could state you have a 0 PPM for the received shipment, but that is not quite the same thing.

0
#180989

Mike Carnell
Participant

Renaud Bompuis,
Why don’t you have them stop shipping defects then the problem of sample size and defects and ppm calculation goes away.
Just my opinion.

0
#180990

GB
Participant

No kidding…
I can’t believe how many Customers fail to act like Customers…I see it every day.   “Nah, let’s not return the part, let’s just rework it here, so we can ship it.”   This goes on and on and on…They have not one clue how much labor and expense they are eating to “just ship it”, and the Supplier thinks their S*** doesn’t stink.

0
#180991

Mike Carnell
Participant

hbgb b^2,
I am having trouble with 8 posts in the string on a SS website and everybody is worried about the calculation and not the defects. I makes a person want to say “Are you serious?”
Regards

0
#180992

Participant

Hmmmmmmmm……..
ppm.
ppm=Pizz Poor Management?

0
#180993

Taylor
Participant

Exactly

0
#180994

GB
Participant

Careful…You swung that bat in Miami and it reverb’d across the country within days…
;-)

0
#180995

Mike Carnell
Participant

At least it wasn’t on video.
Regards

0
#180996

Mike Carnell
Participant

If it isn’t it is close. At least in this case.
Regards

0
#180997

Mikel
Member

Of course the number is useful, you just don’t like the message.

0
#181000

Renaud Bompuis
Participant

Thanks for your answer CT.>Is there no way of matching the error to the order?
There is no systematic way of doing so for defects arising during production or later.Is there no way of matching the error to the order?
There is no systematic way of doing so for defects arising during production or later.Is there no way of matching the error to the order?
There is no systematic way of doing so for defects arising during production or later.>stock reduction
It’s under way. The company is undergoing some changes in the way it manufactures. The high level of stock is generally due to projects that get delayed by the end customer and for which the parts were already purchased.
The factory produces equipment for the railway industry and each project is different.stock reduction
It’s under way. The company is undergoing some changes in the way it manufactures. The high level of stock is generally due to projects that get delayed by the end customer and for which the parts were already purchased.
The factory produces equipment for the railway industry and each project is different.stock reduction
It’s under way. The company is undergoing some changes in the way it manufactures. The high level of stock is generally due to projects that get delayed by the end customer and for which the parts were already purchased.
The factory produces equipment for the railway industry and each project is different.Thanks for the tips.

0
#181008

melvin
Participant

You and your management need to determine what numbers you are really after. Suppliers will incorporate inspection activities to raise the quality of product sent to a customer. This is altering statistical data. Would your company rather have pure data or altered data with fewer supplier defects? If real data is what you want, work out a plan with your supplier. Run a small project to get real data at the supplier before inspection; work to improve their process and to ultimately eliminate the non-value added waste: inspection. A win-win for all.
Counts of defects from the supplier still has value and can be charted. Require suppliers to provide a plan to bring your attribute-data charts into control.

0
#181156

Craig
Participant

I would track monthly, and some cumulative number based on 3 or 6 months.
Line 1 on your chart: Monthly PPM
Line 2 on your chart: Cumulative (total def 3 months / total parts 3 months, expressed in PPM also. This would be a moving window of course).

0
#181160

Renaud Bompuis
Participant

Thanks, we”ll try showing variations of the calculation as your recommended and check what works best for our circumstances.

0
#181174

Scott
Member

Customers Problem Statement: The issue of course is that if we only consider monthly volumes of received/defective parts we get PPMs that are all over the place, for instance.

Objective:  Establish a metric that is of greater value to management.

Assumptions:  (1) You are working to improve supplier quality. (2) An incoming inspection is used to reject or accept the incoming lot.  The lot has been accepted.  (3) The number of parts released monthly into production is a known quantity and equal to N.  (4) The number of defect found each month is D.   Other assumptions are important but these are enough for now.

PPM = (1e6)D/N.

Example:
Step 1.  200 parts are released into production, N=200.  3 defective parts where found D=3 during that month.  Defective PPM = (1E6)(3)/200 = 15,000.
Step 2.  This could be a statistical outlier.  Therefore, you will create a bar graph with two output variables and month as the input variable (excel works very well).  One bar is the monthly PPM.  The other bar is the cumulative PPM.  At the end of the 12th months, the cumulative PPM is the sum of all the defects time 1 million divided by the number of releases for the year.  The cumulative PPM approaches a value for typical performance ability  average PPM.  If you have sufficient historical data the total number of parts released and total defects could be used in a Chi-Square analysis to see if process improvements are statistically significant.

0
#181221

Saurabh Tyagi
Member

Hi,
I have understood the following from your query.
1)You are receiving material in big lots which can be used for two month , three month so on depends upon the production quantity.
2.) You are receivng the defectives in the following month during the production in the previous month lot(Supplier).
3.) But you have not received anything in that month & you dont have the denominator to calculate the PPM.
1.) Use good tracibility system for the parts at supplier end or at your end in order to trace the actual lot for defective parts.
2.) Use the same quantity(Received) as dominator every month in order to calculate the PPM.OR You should collect the defectives parts for one particular supplier lot until the total consumption of that lot & then you should calculate the PPM.
3.) Calculate the PPM in quarter instead of every month for such lots for better understanding.
We are also doing in the same way.
Hope you will understand this in a better way ,
Regards ,
Saurabh Tyagi

0
#181222

Saurabh Tyagi
Member

Hi,
I have understood the following from your query.
1)You are receiving material in big lots which can be used for two month , three month so on depends upon the production quantity.
2.) You are receivng the defectives in the following month during the production in the previous month lot(Supplier).
3.) But you have not received anything in that month & you dont have the denominator to calculate the PPM.
1.) Use good tracibility system for the parts at supplier end or at your end in order to trace the actual lot for defective parts.
2.) Use the same quantity(Received) as dominator every month in order to calculate the PPM.OR You should collect the defectives parts for one particular supplier lot until the total consumption of that lot & then you should calculate the PPM.
3.) Calculate the PPM in quarter instead of every month for such lots for better understanding.
We are also doing in the same way.
Hope you will understand this in a better way ,
Regards ,
Saurabh Tyagi

0
#181298

Renaud Bompuis
Participant

Thanks for your suggestions.The Quality team has settle on a method for now.
I don’t agree with it but they want to try it for a few months.I think in the end your proposal is more interesting.
It’s not feasible for all types of goods since some of them are hard to trace to a particular batch (especially small items that get mixed with previous ones).Thanks again for your input.Renaud

0
#181301

Craig
Participant

Pay no attention to the sigma calculator behind the curtain (Well…not behind the curtain but at the top of this web page). If defects per opportunity are important enough to be in the form of a calculator on this web site, then why can’t someone ask about PPM calculation?
I will give the poster the benefit of the doubt and assume that if they are measuring the supplier’s performance, then they are aimed at seeing improvement. Question 2 will be how to construct a pareto chart!

0
#181303

Ziggyy
Member

PPM by itself is a tyerrible metric to rate anything.  When used as a supplier metric is can vasilate widely from month to month. It is used extensively because it is #1) easy for most people to calculate, 2) Even dummy’s in upper managment can grasp the concept.. higher ppm is bad lower is good.
There are much better metric to use to rate suppliers.
When used in sigma calculation it is a much more delicate function to actually quantify the opportunities for a defect to occur.  Please don’t confuse the two metric.

0
#183757

Carmen Rodriguez
Participant

I would suggest to treat this issue in a two separate ways. One could be the PPM’s result of the defectives detected at Incoming Inspection, and the PPM’s at IPQC, those could be the result of the total used pieces & the defectives found.