One of our customers says we have to establish the " production and quality system to provide a product with zero defects" .We are trying our best to provide zero defect products but has anyone achieved this target. What should be done to ensure zero defect products.

Zero defect sampling does not help as far as I am concerned. What do the forum users feel?

Even if we make the process deliver zero defects, how can we prove it?

If you ever make a process that delivers zero defects you have created a unique process in the history of mankind - well done.

Looks like we found something we can (mostly) agree on, Martin!

IMHO most industrial mfg. or service processes cannot in the real world, over a long period of time, produce zero defects. Maybe, just maybe, if the tolerances of the customer were super-wide open and the product super simple, but how many times does that happen?

A "production and quality system to provide a product with zero defects"? I'd ask the customer how they have decided to achieve that -- ask them and their "wisdom" to educate you.

JMHO

Mike

Over-engineering does take place, but normally not to such an extent as never supplying defects due to tolerances being hugely wide in comparison to process capability.

Most companies would be happy with a process that performs within +/- 3 standard deviations. Leading companies who supply critical products to very stringent customers (such as automotive) would be happy perhaps with +/- 6 standard deviations, and would probably be considered 'World Class'.

Any Phil Crosby enthusiasts out there, or Graduates of his Quality College? I think you realists have distorted the customer's intent. Hopefully you can make defect free product! Now as to some product having a defect, you may be able to talk about the level of quality (e.g., Parts Per Million) but your quality management system should be able to provide defect free product to the customer! Or am I missing something here?

Sorry I think you are missing something.

I am more of a Deming enthusiast, i.e. avoid exhortations in the work place, and dont set unachievable targets !

I don't know, I like all the Quality Gurus (Deming, Juran, and even Crosby) they have good principles to build a Quality Management System. But saying that you can not create a system where you are competent that only so many defects would be present, is someone I would not likely choose as a supplier.

As a "Shy Browser" this will be it from me!

Douglas

Admittedly I do not know the full content of Crosby's Zero Defects system, however by the title I have always imagined that it meant producing totally defect free product - I dont believe that can be done. As for having a consistent output of 'nearly' defect free then yes that can be done, and is our never ending goal to improve on.

But is zero defects achievable ?

Have you witnessed this ??

Given the recent developments in quality as a partnership and continual improvement, the concept of zero defects might be given a different look. If you consider the definitions of defect and nonconformity it could possibly be applied.

Defect: - a departure of a quality characteristic from its intended level or state that occurs with a severity sufficient to cause an associated product or service to not satisfy intended, normal or reasonable foreseeable usage requirements. (Armand V. Feigenbaum – Total Quality Control 3rd Edition Revised)

Nonconformity: - a departure of a quality characteristic from its intended level or state that occurs with a severity sufficient to cause an associated product or service to not meet a specification requirement. (Armand V. Feigenbaum – Total Quality Control 3rd Edition Revised)

The concept of “Zero Defects” now becomes a commitment to not supply any product which would cause an associated product not to satisfy intended, normal or reasonable foreseeable usage requirements. This is certainly achievable within the statistical boundaries set by a stable process. A nonconformance to this commitment can be classed as a “Gross Nonconformance”.

Now quality becomes a true partnership. Assuming a capable and stable process, the supplier has committed to not supplying a “Gross Nonconformance” while continuing to improve the process by reducing minor nonconformance. The customer, now confident of “Zero Defects”, accepts and uses product supplied under a normal capable and stable process knowing any minor nonconformance which may statistically be present will not be detrimental to his product.

Obviously, in addition to the supplier/customer partnership, the definition of “Gross Nonconformance” is critical to satisfactory implementation of this “Zero Defects” concept. It could be stated that any nonconformance undetectable to the customer through subsequent testing or use could be called “minor”. Anything the customer could detect would then be classed “Gross”. This may not, in all circumstances, consider the “usage” requirement identified earlier.

The preferred method would be to define “Gross” at the APQP stage. Each part controlled and monitored with its own criteria. Minor nonconformance would be brought to the attention of the customer and a mutual agreement to “usability” would be made.

Given the right elements of partnership, planning, communication and process stability, it would seem possible to reach a plateau of “Zero Defects”.

Just my opinion but it is what I am preaching to our company.

Dave

Maybe I'm looking at this too simplistically (is that a word?)

The customer has requested that you provide them with a product with zero defects. I read this to mean that the customer does not want to see any defects from you. This doesn't mean that you cannot produce defective product, it just means that you must catch it before it reaches their dock.

Does this make sense?? A production manager's goal is always to produce zero defects, but the ultimate goal of a company is to supply product with zero defects. Two different scenarios, although achieving the first part will make the second automatic.

Just a thought.

Here's something I ran across while surfing last night. While I didn't read it thoroughly, it appears to "coincide" with Dave's response.

http://www.hands-ltd.demon.co.uk/zerodef.htm (http://)

Maybe it will help some.

Bill

I've never seen a zero defects process either.

Just a random thought on a Friday morning... if one has a true zero defects process, then does that mean one cannot show continual improvement? Wouldn't that be a scream on an audit finding?

Now back to work.

As a "Shy Browser" this will be it from me!
Whoa there Douglas! We won't let you off this easy! :mad: Your comments are just as valid as anyone else’s!

but your quality management system should be able to provide defect free product to the customer! There is a difference between manufacturing a defect-free product and providing one to the customer! The customer wants no defects. You can produce all the defective products you desire, just don't send any to me!

Our goal should always be to have zero defects. It might not always be possible, but we can get really close. And we should get even closer to shipping defect-free parts.

One of my clients makes small, inexpensive vinyl parts. One of their customers wants zero defects. The process is not capable of making zero defects. So my client approached the customer and let them know that if the customer wanted zero defects there were two possibilities. First, they could make the parts using a different process that would triple the price, or the customer could pay for 100% inspection that would also add greatly to the cost. The customer decided the current level of defects was satisfactory. (Apparently the customer contacted other manufacturers who basically gave them the same story)

Spaceman

......and lead to mass redundancy in the quality function !

QUOTE: "If we are talking about product defects, then a process that produces zero defects might still be able to be improved in efficiency terms."

Just ask any customer - once you have cleared the 'zero defects produced' bar, then you need to do it faster and cheaper!!!:frust:

If zero defects are possible?? it depends how many parts you are producing and if you have processcontrol.
In theory process controll of 1,67 will give you if I remember right 3-6 defects per million, that is according to the automotive industry OK, BUT, at the same time they want 0 defects. They are saying 2 things that are going against each other.

Even if you have 100% inspection it will give you a ppm of 300.
Can you have 0 defects? YES but this will cost the customer and nobody is willing to pay for it, instead the want cost-reduction and better quality, somewhere that does not add up.

Here is what they do: you may deliver parts to them for a year and have no defects, then at the end of the year you may have just 1 defect. You will still get that letter that you are not good enough and you will pay for all the expenses, and they will leave you if you don't get better. My suggestion. Find other customers that are not in the automotive industry.

I have been to the Crosby Manager's training, their SQC training, and trained the beginning class more times that I want to count. This sounds to me like a misappropriation of a catch phrase, and a gross misuse, possibly.

Crosby's "Zero Defects" DOES NOT mean that we don't produce defects. The key is the reaction we have when they do happen. We fix the problem, analyze the problem (root cause) and we make the problem go away forever. In other words, it is an ATTITUDE!!! Not a specification.

If they really think they are really never going to have a defect, then they are either really nuts or they arenot doing a good job defining requirements (Crosby: Quality=conformance to requirements).

I suspect they heard the phrase somewhere and decided it would be a good idea. Mention Crosby, and if you get a blank look, duck.

I think zero defects is acievable, especially if you look at shipments to customers, for instance, one part, five complaints in a month, even with full 100% + 100% inspection still some came through as the defect was extremely small, cost last year, some $35K, found some home/garden wizzard to make some wierd looking, sensored inspection machine, cost $10K, defects found (@ customer) 0 in 10 months, (1 Million parts) efficiency up 10% as inspection is no longer bottle neck, who says it can't be done.

Actually last 5 months one of our facilities has been running zero defects, ....better knock the bottom of my desk.

Our major customer sends us every part they find bad as they assemble it.
I know we need to use the 80/20 rule, however my production manager claims the largest defects which is cracks is caused by material. But material is hard and costly to get so we have to live with it and thing of another way.
We are currently looking at montoring the loads of certain stations on the press, to see if it would detect cracks. We already tried a photo eye cell in a die that would stop the machine if there was any light. But the phot cell was always breaking.
Anyway there are many parts with simialr defects sent at this time it seems weekly. I am interested in how I can organize and measure our improvment or failure to approve to present at management review. Of course , i would love to hear from anyone who has coorected cracks in drawign cold roll steel.

Our major customer sends us every part they find bad as they assemble it.
. . .
Of course , i would love to hear from anyone who has coorected cracks in drawign cold roll steel.This has been a very interesting thread to read.
Dave Scott's post http://69.93.111.34/Forums/showpost.php?p=39952&postcount=9 closely approximates my understanding of "zero defects."

Back in the early 90's, my company paid for two full page (2 sides) 4 color inserts in two separate volumes of the Thomas Register, where we proudly proclaimed "Zero Defects!" Funny thing was nobody ever challenged us on that claim (although everyone here knows it was and is categorically impossible to deliver absolutely zero defects.)

What we did do was work hand in glove with every customer to eliminate as many of the "we said, they said" stumbling blocks before we ever cut a piece of metal.

Once we began cutting metal, we kept up a steady dialog with each customer throughout the production run. Sure, we produced some nonconformities, but we were proactive when we discovered them and quick to react, research root cause, and correct when our customer discovered them.

In short, it was how we responded to a nonconformity which kept our customers coming back, not whether we provided "zero defects."

yes, I agree. Responding in a timely manner in both a corrective and preventive way is the best approach. I am speaking about measuring our performance. However, there is no doubt that the cultue has to be there for an effectve corrective action process. At least two of our major problems are going to take a long term development plan, meanwhile we are absorbing sort costs. Sorting 100% is not as effective to prevent these bad parts from reaching the cusomter as I would like

yes, I agree. Responding in a timely manner in both a corrective and preventive way is the best approach. I am speaking about measuring our performance. However, there is no doubt that the cultue has to be there for an effectve corrective action process. At least two of our major problems are going to take a long term development plan, meanwhile we are absorbing sort costs. Sorting 100% is not as effective to prevent these bad parts from reaching the cusomter as I would likeI'm painfully aware of the costs of sorting.

One of my suppliers had a major "disconnect" in a process and ended up with 100,000 suspect pieces. They were sure the only way to sort to eliminate most of the nonconforming parts would have been to build a special gage to mimic the mating part, then run each and every piece through the gage. Alternate methods they thought of were even more costly and labor-intensive.

I became aware of the situation when the supplier missed a partial delivery of 5,000 pieces. The fluster and bluster of conflicting excuses from two different folks at the supplier induced me to have one of our engineers who lived near the supplier take a short detour to check out the situation first-hand.

When my engineer phoned from the supplier's shop, he was laughing so hard, I could barely understand him.

He said, "You won't believe what these guys are doing! They screwed up 100,000 parts that they make for a dollar each and sell to us for a dollar fifty. Now, they are planning to spend $3,000 to build a gage so they can sort the parts at a rate of one per minute. They were too embarrassed to ask us for mating parts to use for gages. They know why their machine screwed up and can knock out 5,000 good parts today, but they didn't want to spend $1,000 for bar stock to make the parts! Should I explain the economics to them?"

I said, "Put the plant manager on the phone. I'll talk to him."

We got our 5,000 good pieces that day. We sent over 25 mating parts to use as sorting gages ($5 each mating part and the net time to check each piece was only 20 seconds) and the supplier learned one lesson he'll never forget - When in doubt, tell the truth! We also helped our supplier learn about SPC and in-process inspection for future work. All in all, our supplier increased his profits as a result of the SNAFU. He learned to look outside his firm for answers to avoid "can't see forest for the trees" syndrome and he learned the value of maintaining customer dialog.

Bottom line:
Supplier-customer partnerships really can work, but only if both parties contribute to the effort. Chief goal - eliminate fictions like ZERO DEFECTS.

One other thing: I LOVE the way most engineers think - they usually look for an "elegant" solution (one that is simple and inexpensive) because they believe in Occam's Razor. Our engineer was a great example.
OCCAM'S RAZOR
named after William of Occam. Given a choice between two explanations, choose the simplest -- the explanation which requires the fewest assumptions.

Wes, Good post. I do have a couple of questions.
Are these automotive parts supplied to an OEM or Tier I? If so, how do you address the sorting process using mating parts with your customer?
Even though using mating parts is an efficient method there is still the possibility of accepting product that is out of tolerance.

Wes, Good post. I do have a couple of questions.
Are these automotive parts supplied to an OEM or Tier I? If so, how do you address the sorting process using mating parts with your customer?
Even though using mating parts is an efficient method there is still the possibility of accepting product that is out of tolerance.Excellent question!
In our case, we were the OEM. Our engineer was, in fact, the design engineer responsible for the product. If we had been supplying this product as part of an assembly to another OEM, we would have immediately involved the customer in the discussion/decision.

Regardless, we were the next link in the chain and responsible for all the preceding links (from the metallurgical qualities of the steel alloy to dimensions of the piece part.) If we had been a middle link, rather than "design responsible," we most likely would have sought a written waiver for the change in process of the Quality procedures.