PPM (Parts Per Million) Measurements - Parts vs. Opportunities

[ckochen]

Our company is trying to determine the best way to calculate ppm for distributors of manufactured electronic parts. If a supplier sends us a reel of resistors that has a qty of 100,000 do we count this as 100,000 defects on their ppm or 1 since they only commited one picking error.
This also beings up the issue of calculating an overall ppm when the parts come in with different units of measure ie. each, ft, litres. Can ppm values based on different units be averged togther. ie ppm of each, ppm for length.

 

[Brizilla]

Re: PPM measurements - parts versus opportunities

I'm not sure if this would apply to you, but we apply differentiating units into PPM by calculating $defective/$received.

 

[Phil Fields]

We count every part, even if is one box of 2,000 bolts. Just because it is 1 box, I am still down and can not produce product. If we buy in "feet or pounds" we will reject in “feet or pounds”. I try to keep it all one to one.

When I do this with a distributor I explain my method of determining the PPM. I also let them know that they are in the distribution business and should be experts in warehousing. They do no value add to the product, they purchase, store, pick, ship product.

I did this recently with a hardware supplier that was shipping in the incorrect product(s) for three month (different product). Six month later I have not had an issue.

At the end of the month I evaluate total number of parts rejected, and the total number of nonconformance reports (NMR) issued per supplier. A supplier my have 2,000 parts rejected with one NMR. Or a supplier may have 50 parts rejected with 25 NMRs. The supplier with more NMRs is costing more time and money than the supplier with one NMR. So I do not look at number of parts alone to determine supplier quality.

 

[Helmut Jilling]

Our company is trying to determine the best way to calculate ppm for distributors of manufactured electronic parts. If a supplier sends us a reel of resistors that has a qty of 100,000 do we count this as 100,000 defects on their ppm or 1 since they only commited one picking error.
This also beings up the issue of calculating an overall ppm when the parts come in with different units of measure ie. each, ft, litres. Can ppm values based on different units be averged togther. ie ppm of each, ppm for length.

The common defintion for PPM is Defective parts per million. So, yes, every defective part would count. Non-defective parts should not be factored in because they were not defective.

Defective ppmo is a little different. The theory here is that a complicated system (ie: transmission) has many more opportunities for defects than a simple metal bracket. So, measuring the number of defective UNITS could be a pretty high number, based on the complexity. There is some validity to this, but most customers track it based on units, not opportunities. Internally, some companies base it on opportunities, especially if they are running a six sigma program.

So, here's how it would look.

Defects per million units would be based on defects / 1,000,000 parts. 1 defect in a million parts would be expressed as 1 ppm.

Defects per million opportunities would be based on number of parts x number of potential defects per unit (opportunities). So, 1 defect in 10,000 parts x 100 potential defects per unit (opportunities) would be 1 ppmo.

 

[Miner]

Our company is trying to determine the best way to calculate ppm for distributors of manufactured electronic parts. If a supplier sends us a reel of resistors that has a qty of 100,000 do we count this as 100,000 defects on their ppm or 1 since they only commited one picking error.

How many nonconforming parts did you receive, 1 or 100,000?

You probably ordered 100,000 resistors, not 1 reel. Count the defectives in the same units as you ordered. 1 defective reel / 1 reel ordered, or 100,000 defective resistors / 100,000 resistors ordered. You get the same answer either way.

This also beings up the issue of calculating an overall ppm when the parts come in with different units of measure ie. each, ft, litres. Can ppm values based on different units be averged togther. ie ppm of each, ppm for length.

If it works for you, yes. However, You will probably run into problems with the relative weighting of the different units.

If your lengths were measured in meters or in millimeters, the relative weighting of the second would be 1000 times greater than the first. If you can apply some rational rule such as one 12-foot length of bar stock is equivalent to 12 one-foot shafts and weight it accordingly, it might work.