Measuring Quality Performance of a Jobshop production 1

[JosephERG]

Hi Everyone,

Good Day!!!

I would like to ask advise from you guys (especially from the "jobshop people") regarding the best method for measuring quality performance of a jobshop production.

As of now, we are using the yield, %rejects, abnormal
lots, and returned lots as our metrics for quality performance. I am not really sure if these are the correct
metrics because sometimes we have good numbers but we still
receive customer complaints.

Should we use CpK, DPPM or SPC instead? We have considered using them also but we had difficulty because of the high variation and low volume (HVLV)of our products.

Hoping for some positive responses from you guys.......

 

[tomwalz]

Dr. Deming says that quality is exactly what the customer says it is.

What are they complaining about? Find that and measure it. Make sure you find out why they are really unhappy. If they fought with their wife that morning your parts are more likely to be rejected or, at least, be the subject of complaint.

Maybe do a "Dog and Pony" show meaning a small circus. Take the sales rep, an engineer, somebody from QC, a machinist or two to the customer and listen to everything anyone cares to tell you. Maybe some of them are a slightly different color. Maybe they don’t like one delivery driver. Maybe they don’t like the protective coating you use.

Four rules:
1. It is hard to measure something unless you know what it is you are supposed to measure.
2. Communication is always imperfect. With perfect drawings in an agreed upon format and a great deal of discussion you may hit a communication level that is 90% accurate.
3. Many times people know something is wrong but don’t have the education or vocabulary to accurately describe it. I see this a great deal when there is a dispute between engineers and production people. The production people know something is wrong but they don’t know how to talk to engineers so that engineers understand. Engineers are often all to often ignorant of the actual production world. Many engineers think of metal as homogenous with uniform properties throughout. Darned few machinists do.
4. They are people. I am sitting here just furious because my expensive web site consultant has our web site all screwed up. I told my staff and they are leaving me alone. I will walk the dog down to the corner and get a cup of coffee and I will be rational again. (or as close as anyone ever gets.) However now is not a good time for anyone to bother me. If I were to inspect parts now I suspect I would find more cause for reject than if I were rational.

Make sure they like you and know that you care. Show up with doughnuts or pizza. Explain what you are doing to solve their concerns.

Or, if they are really jerks, drop them and run. You will probably have trouble getting paid from jerks, anyway.

O.K. I am better now. Thank for letting me rant. Time to go tell the staff I am human again then a dog walk and coffee.


Thomas J. Walz
Carbide Processors, Inc.

http://www.carbideprocessors.com

Good engineering starts with a Grainger Catalog.

 

[tygerdawg]

I was gonna say something similar but Tom's rant covered it mostly.

You say your "numbers are good" but they are rejecting parts. That doesn't add up. You must be measuring the wrong things, and not finding out what the customer wants. Or, as might be the case, what the customer needs. Sometimes they don't know.

I once watched a new product start up in my factory. Our company was invited to steal the business from another supplier who "was gouging" the customer. Manufacturing the part looked straightforward, and the pricing was 1/3 of the original supplier. Slam Dunk. Everything was set up, run by the numbers, met the customer drawing specs. Parts were rejected by the truckload. After many meetings with deep statistical analysis on part features (we utilized Shainin Six Sigma methods) it was discovered that the original supplier deburred machined corners with (expensive) Method1. Our company decided to use (cheap) Method2 for that process step. Both companies met the design print. It turned out that Method1 produced a sharper corner, which produced a longer sealing land on the face, and therefore did not leak. Our parts leaked because Method2 rounded the corners more, but the parts still met the design prints.

The customer never understood the design nuances of their own parts. All they saw was Company1 parts worked, Company2 parts failed.

Gawd, it sure got ugly after that.

Your description is not that clear and leaves a lot open for interpretation. But if you are making the parts to the print, but the customer is rejecting the parts, then it seems that the problem rests with the customer.

TygerDawg
Blue Technik LLC
Virtuoso Robotics Engineering

http://www.bluetechnik.com

 

[arunmrao]

tygerdawg,

This conforms my statement,"drawing does not say it all". I go the extra mile of visiting the customer,meeting different shift foremen/millwright mechanics,understanding,the fitment,usage and problems faced from the competition product,before I embark on producing my part. It has worked well,but not always resulted in getting revenues ( for business reasons).

 

[KENAT]

Hmm, interesting views on the value of a drawing. I'm inclined to think a well done drawing (or equivalent MBD) lays the foundation for being able to get the parts you need.

So many times I've had phone calls from vendors with questions or received parts from vendors that weren't good where the clearly hadn't paid attention to the drawing. The amount of times I've been asked questions and my answer has been 'see note 6' or 'look at view A-A' or the like. This especially happens for parts we use drawings to define but give a courtesy/reference copy of the model for them to use in CAM to speed things up. Sometimes it's as if they just throw the model into the system without paying any attention to the drawing and ship us whatever it churns out.

Hell, just yesterday I got CC on a drawing from a vendor asking if they really needed to do the engraving called out on a drawing. I had to stop my self from replying along the lines of 'if it wasn't needed why would it be on the drawing' - and this isn't a new drawing that might be prone to error.

So my first suggestion would be to make sure you read & understand the drawings. If the drawing is ambiguous - as all too many are - ask for clarification before you start machining - or even better before you give a quote/accept the business.

Of course, this is an engineers point of view - purchasing may not like the vendor that asks too many questions rather than just giving them a quick quote. They may not see the benefit in getting an accurate quote up front rather than all the tooing and throwing, schedule slips etc. that happen when they didn't fully understand the task.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[KENAT]

Sorry, I got CC on an email, not a drawing. - doh

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[JosephERG]

Thanks guys for the valuable insights.....

I agree with what tygerdawg said about the mismatch between our "numbers" and our actual performance. We failed to recognize that what was "good numbers" for us may not necessarily mean "good" for our customers.

Actually, I am focusing more on preventing or putting a stop (for good!!) to our customer complaints. We did some analysis on the causes and we found some commonality:
(1) Knowledge & Skills mismatch ? actual knowledge & skills of our workers do not match the requirements or complexity of their jobs.
(2) Process capability ? our processes are manually performed, so there is high variation.
(3) Communication ? this involves (a) dissemination of new information and (b) interpretation of customer drawings.

Regarding customer drawings, we don't have the luxury of negotiating with our customers for an agreed upon format (just like what tomwalz suggested) - although we've already opened this up to management a long time ago. Therefore, we are stuck with accepting drawings that have various formats and interpreting each of them.

We have identified some actions to counter the 3 causes we have identified. I would like to know your ideas also on this matter. Appreciate it if you can share some.

Thanks......

 

[KENAT]

Well, do your staff have any education in how to properly read an engineering drawings that's been prepared to one of the recognized industry standards?

Of course not everyone closely follow such standards but understanding general drawing conventions etc will help.

Also, if some customer drawings have GD&T you need someone that can interpret it, both for machining and inspection.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[tomwalz]

I am not sure that eliminating customers’ complaints completely and for good is the best approach. I wish it were.

Your writing seems to indicate that you are much more intelligent and better educated than most. You also seem to believe that communication is the burden of the communicator.

Unfortunately much of the world is different to very different. I believe that communication is the burden of the communicator but many do not. Also many will do a quick drawing expecting you to call if you have questions.

We decided years ago to take drawings on napkins and similar. If we need to we re-draw them. It seems that reading a drawing is easier than creating one. If we do a recognized format drawing and send it to them for approval they can generally read it. It is much like the old trick of repeating something back in different words.

I love sample parts and digital cameras. We will do sample parts whenever possible. A digital camera picture of a mock up can be hugely helpful.

A concept part in foam or wax or wood can be tremendously helpful. You have a lot of manual machinists and somebody ought to be able to get one done pretty rapidly. Send a picture. We have even done them in paper.

I have red fingernail polish sitting on mu shelf I use for identifying a particular side. Not knowing which side is up is not as funny as you might think.

As for your manual machinists and precision, repeatability, etc. I would investigate, stops, gauges, etc. There is a huge variety of tools available to solve or greatly help with this issue. Ask some of your guys and see what they say.

If it is a machining issue I would post on a web site dedicated to practical machining issues.

Training customers is a long, slow process. Even then it sill takes a lot of questions to get things correct. I really do believe that even the best drawing communicates only about 90% effectively. Sometimes I am wrong but it is a good place to start.

Tom


Thomas J. Walz
Carbide Processors, Inc.

http://www.carbideprocessors.com

Good engineering starts with a Grainger Catalog.

 

[JosephERG]

Thanks guys for the advise.

By the way tomwalz, is it OK if you share those machining websites you mentioned? I'd like to check it out.

 

[tomwalz]

I don't know but I'll try.

practicalmachinist.com

Huge store of knowledge. Real world guys. You will definitley get a plant floor viewpoint.

One of the best things that happened to me was having to work my way through college on a plant floor.

Tom

Thomas J. Walz
Carbide Processors, Inc.

http://www.carbideprocessors.com

Good engineering starts with a Grainger Catalog.

 

[JosephERG]

Thanks tomwalz......

How I wish I had the same opportunity as you had.

By the way, I'd like to know (if you don't mind) if your'e also into "process capability" (CpK, Cpu) or SPC in your shop floor.

We haven't really applied any statistical controls on our processes (especially machining). But there is this customer of ours who required us to do so just recently. They want to know our CpK.

Do you think statistical process controls are possible in a jobshop? I have read some articles about this matter and they're not very encouraging.

Any advise? Thank you very much.