I thought the following was interesting.

-------Snip-------

From: Hakan Sodersved
To: DEN Disc List
Date: Tue, 12 Jan 1999 13:07:32 +0100
Subj: Going Crosby vrs Going Deming

I see Crosby as a natural 1st step to quality from the curriculums that trade and technical universities had during the 60:ies and 70:ies, with some exceptions.

At the Technical University I spent my 5 years in 1966 to 1971 we had a thin 0.7 cm book called The Psychology of Worklife, and we disliked that subject and correspondingly the lecturer. This was the only touch of Profound Knowledge/psychology we were offered. I wonder how many youngsters 18-25 are interested in that type of psychological learning. It sure was much different from the method of Mt. Edgcumbe School in Sitka, Alaska in the 90:ies.

The knowledge given in economics was not far better at the technical university. Even at the end of the 80:ies in my industrial career I was regarded a specialist being able to calculate costs of products in mass production. Cost of quality was not included then either, it could not simply be measured due lack of quality cost data and inaccuracy and bias of sampling in the economical systems. A natural step would have been to go for Crosby, because more and more people were interested in economical figures of their products and processes. But due to a strong central quality institute, Deming was dominating.

To introduce something in a large organisation you must have broad support for the new thing. It is easier for a broad mass with the Crosby concept than the Deming philosophy. But the journey will take a longer time.

But for me personally, when I "detected Deming" in 1989, it was an excellent timing. It was far more offensive and complete than the fresh ISO9000. It gave light to all my questions about the Japanese mystery of success. It also gave support for many results of my personal research "on the production floor". My big problem in the beginning of the 80:ies was a dramatic unacceptable increase in production costs, year by year. This kind of "dynamite knowledge" is difficult to deploy if you have an offside position in an organisation. It is a rare chance for an internal person with appropriate authority and profound knowledge to get the opportunity to gather facts and data like Dr. Deming has been able to during his trips.

The Deming philosophy deals with senses of the whole brain - not just logistics and economics. You include ethics, empathy, feelings, beliefs, values, logic, system frequencies of interaction. The Symphony Orchestra is a very good socio-physical-technical model (I exclude the musical styles and preferences). Crosby also has a lot of ethical values in his approach and he often speaks the language of the top.

I always recommend Deming-what else-but I try to understand why the Crosby step might be a necessary as an intermediate step. The two routes may have orders of difference in quality magnitude. Who could measure?

http://www.expira.se

-------End Snip-------

Thoughts, Anyone?

Regards,
Don

The Deming philosophy deals with senses of the whole brain - not just logistics and economics. You include ethics, empathy, feelings, beliefs, values, logic, system frequencies of interaction. The Symphony Orchestra is a very good socio-physical-technical model (I exclude the musical styles and preferences). Crosby also has a lot of ethical values in his approach and he often speaks the language of the top.
I think this is the important part of Deming - which is part why I hate the Quality word. The approach has to be holistic and it has to be appropriate for the type of company and processes.

Don,

Good posting. No secret here, I'm a Deming disciple!

Good posting Don! I could hear Deming's coarse monotone voice uttering the words "highway down the tubes". Deming's focus was on controling the inputs for system optimization. I guess Zero Defect output could be considered an ancillary benefit and not the goal. The goal is to reduce system variation for system optimization.

Deming's focus was on controlling the inputs for system optimization. I guess Zero Defect output could be considered an ancillary benefit and not the goal. The goal is to reduce system variation for system optimization.

That is the element lost on some folks. Spending your time on the outputs is not the key. Controlling inputs, through continuous monitoring and improvement is the key to process improvement.

Regards,
Don


[This message has been edited by Don Winton (edited 01-15-99).]

Marc,

Agreed.

More

-------Snip-------

In a message dated 1/14/99 5:00:03 AM, info@expira.se wrote:

"I always recommend Deming-what else-but I try to understand why the Crosby step might be a necessary as an intermediate step. The two routes may have orders of difference in quality magnitude. Who could measure?"

Deming could and did *measure* when he called the Crosby concept of Zero Defects a *highway down the tubes*. At first I thought he was kidding, but he assured me he wasn't. It was Zero-Variation-of-the-System-Operating-Components we should be after, not Zero Defects I learned.

By the way, when I was General Manager at FPL QualTec, many of our clients were Crosby followers who had hit the wall, had spent hundreds of thousands of dollars pursuing quality improvement, and were floundering. For them, the Deming method definitely proved to be the better way.

Frank Voehl (FVoehl@aol.com)

-------End Snip-------

Regards,
Don

[This message has been edited by Don Winton (edited 01-15-99).]

So much focus on the bottom line. When the bottom line isn't what we 'expect', then we begin to look into 'why' that is so. How did we get here?

A quick review of a process may show that a predicted output (most likely arbitrarily selected and w/o statistical significance) did not meet our expectations. Should there be any surprise? A quick review of the inputs may show that the process was never capable of delivering the arbitrary target selected. There should be no surprise. Everyday, demands are placed on systems everywhere that are not capable of delivering the goods. Does this stop us? No, but this can lead to frustration, confusion, or worse. Have we done this? I am personally guilty. Deming's philosopy on system optimization teaches us to avoid these pitfalls. He did not sell you that you would produce zero defects, he sold you that you could. Understand the process inputs, control the output and achieve predictability.

For me, I feel Crosby's 'Zero Defects' concept uses zero defects as a starting point. This creates the arbitrary target in my mind. No plan on how to get there, just get there. Dangerous in my opinion. I know some out there will point out that this is not what Crosby means. I would agree with this. Still, 'Zero Defects' connotes to me a 'suggestion without a plan' and not a solution. No right or wrong about it.

But when I hear 'System Optimization', this denotes to me exactly what must be done. P-D-C-A using real data and using this to predict system output. Refine the system under continuous improvement and approach zero defect output as an end point.

Back to the group...

Kevin,

Good points all around to which I have no arguments.

A quick review of a process may show that a predicted output (most likely arbitrarily selected and w/o statistical significance) did not meet our expectations. A quick review of the inputs may show that the process was never capable of delivering the arbitrary target selected.

This is a point I try to drive home over and over again. Outputs (far end) are not important. It is the inputs (near end) that are critical. Using available data, resources and tools control the inputs. When the inputs are controlled, the path to systems improvement (Low nonconformances) is paved with gold. The outputs should be monitored for progress, but eventually it will be found, through proper techniques, that the outputs will be predictable with changes to the inputs. A particular company I know spent hundreds of thousands of dollars on inspection and test (automation, computer controlled, etc.). State of the art? Yes. Impressive for visiting dignitaries? Yes. Useful? Who Knows. A visit to the fabrication facility revealed a complete lack of control of the inputs. Zero control on oven times. Zero control on sealing methodologies. Virtually zero control on any of the key input variables. When I asked about this, I was reminded of deer caught in headlights. They did not have a clue.

Regards,
Don

Don,

I agree with the blank stares statement you made. All to often the case.

What do you mean? I spent thousands of dollars on inspection devices! My quality is good! Right?

To borrow a line from Deming; "Wrong!". To borrow another line from Deming "How could they know?" is a classic case of pointing out that we embrace the wrong idea. We are sold that Quality is inspection. Case in point, several suppliers have Quality Manuals that reflect nothing more than an inspection program. When submitted for review, I ask if they have plans to document the Quality System. The blank stares and the "what do you mean?" statements are common. They had believed that they had done so with their Quality Manual. They do not know the concept of systems to prevent nonconformance, they practice systems to detect nonconformance. This is because they have adopted the existing philosophy (the Western Management Philosophy) and the 'Caretaker role' (they may have even borrowed someone else's QM to copy). What worked in the past, or for others, will work us in the future. Right? Folks select these ideas because they are taught them from the start (a monopolistic idea). No other solutions are taught so we continue to stumble along. What is worse is that we accept mediocre results, sometimes even wave them about as achievements. Is anyone to blame? Pointless I think. I feel we were advancing along in the industrial age and just became to comfortable (we forgot to keep growing). Now our slackness in forward progress is catching up with us. Enough with the spilt milk, let's clean up and move ahead (no finger pointing)! (did you notice that I worked Deming's and Crosby's ideas into the post?)

Embrace the philosophy of optimization and prevention. These ideas can go a long way to achieving organizational immortality. Back to the group...

For those interested, you can try this.

I had forgotten about this paper. It is an interesting read.

Regards,
Don

March 2004 - The link is dead but the paper is attached in a post below.

Good thread.

I recently had a quality issue at a customer( heaven forbid, but it happened). This happened at an international customer site. We happen to have a subsidiary plant of ours, about two miles away..we sent personnel from there to investigate. As it turned out, there was a small amount of product that had been assemble incorrectly. Long story short, the design guy at our plant there, got a lecture form the manufacturing engineers at the customer, on the IMPORTANCE of inspection at the end of our assembly lines. He relayed this information to me.......which launched me on a tirade of pro-deming, anti-inspection philosophy. I basically have cut the inspection processes (and the costs associated with it), by about 75% since my arrival at my current job. We have re-directed those energies, (and funds) into poka-yoke techniques in design and manufacture. Anyway, I ended my spiel by challenging to compare our costs of inspection, vs. customer rejects, (before my arrival), to the costs of prevention, and the subsequent reduction of customer problems, (after my arrival). By the way, the particular product which was a problem, had NO poka-yoke in the process, which was a failure of the system.

Deming???? You bet. Works for me. Prevention is the only thing that makes sense. It is disheartening to see, (again) that the old Taylor stuff is still alive and well in our midst.

Bryon

Bryon

the design guy at our plant there, got a lecture from the manufacturing engineers at the customer, on the IMPORTANCE of inspection at the end of our assembly lines.

I get this all the time. When it becomes totally irritating, I usually give the “giddy girl” test. You know, a story where you count the “g’s” with those being “defects.” After completion, usually most only get about 85% correct. Then I drop this:

If your inspection techniques are 85% effective, that is normally considered normal, right? Most replies, “Yea, but it could be better.” Then I ask, “Well, what is acceptable?” I then get something to the effect that 90-95% (any 100% answers are dismissed for reasons I have covered in other threads) would be considered acceptable. Then I drop this:

OK. We have decided that 95% effectiveness is acceptable. Under the assumption that 1000 items are inspected at 95% effectiveness, what is the probability that a product may ship with a defect? Blank stares. Then I demonstrate:

0.95^1000 = 5.29e-23. In other words, in 1000 opportunities, it is a virtual certainty that some product will ship with some nonconformity.

I readily admit that the devil is in the details. The above is not precisely correct, but it makes an impression.

Prevention is the only thing that makes sense.

Agreed. By implementing prevention, it raises the probability through redundancy. When effective prevention is implemented, effective inspection is not required.

Regards,
Don

Don,

Nobody looking for 100% effectiveness. I wonder why? What was it that Deming said? "Living in a world of defective product, as if it were necessary" (or something close). Inspection has it points, but catching product at the end of the line is "too late"! Nice diiscovery learning techniques applied (giddy girl). Nothing more concrete than learning by practical experience.

Inspection has it points, but catching product at the end of the line is "too late"!

Yea, Deming put is something like this: "Evacuating the hotel after the fire has started is not improvement." I am probably paraphrasing, but the point is still there. The hotel still burns.

Nice discovery learning techniques applied (giddy girl).

You would be surprised (maybe not) at the frustration managers exhibit when given this test. Especially when I am screaming over their shoulder "why ain't you done yet!" and interupting every few minutes asking what is the status of so-and-so project. (simulating a typical production environment).

Regards,
Don

Terrific thread. You guys said it all.

May I propose a favorite tool of mine that identifies INPUTS and facilitates control of causes - the FMEA. If I was to identify my biggest hurdle in managing quality, it would be bringing folks on board regarding the differences between causes and effects. A well done FMEA can help in this effort. Until a couple of years ago, we brought on new projects using FMEA development - flow diagram first, inputs/outputs, sources of variation, CAUSES and effects - as the most important single method of identifying potential and then "controlling" (eliminating when possible) causes. We also used past FMEA's as a tool in continuous improvement / problem solving efforts. We always started out with the idea that defects were unacceptable.


Can someone email me the "giddy girl" thing? I used to have a "fowl farmer" thing that I used to use in my classes, but I seem to have misplaced it. Thanks again.

Batman,

May I propose a favorite tool of mine that identifies INPUTS and facilitates control of causes - the FMEA.

Agreed, but there are other methods, Slater’s book, Integrated Process Management, identifies these. That is not to say whether one is better or superior over the other, just there is an alternative to the standard FMEA method. Your method of identifying INPUTS rather than OUTPUTS is an excellent method of systems improvement.

Can someone email me the "giddy girl" thing?

I have a Word 6.0 copy at work. Will send Monday. BTW, I have other versions of this test that I will attach as well. Some are more effective that others when getting the point across.

I particularly like interrupting the test takers asking “look into this” or “what is the status of” type of stuff while test taking. Simulates what inspectors experience and gives MANAGERS what they are expecting of their inspectors.

Regards,
Don

Don,

I have always enjoyed the discovery based learning approach. The realizations that are achieved outperform the* mere teacher-desk-pupil paradigm of learning (at least for me). The interuptions you insert into your practical learning methods epitomize real case settings. This test illustrates the paradigm of "Zero Defects" very well.

Batman,

The FMEA is a great tool for Risk Analysis. Big in the automotive world, lesser known in the Medical Device world (Essential Requirements). Either way, each serves the same purpose, elimination of potential risks (Prevention). It appears that your folks are utilizing several of the Statistical Tools. Good stuff! Statistical Thinking was a large part of Deming's world.

I have always enjoyed the discovery based learning approach.

Yea, you should see the stuff I get when I do the ‘Match’ test. The concept is communication. I ask the class to instruct me, without any prior knowledge whatsoever, to instruct me, in a written instruction, how to light a match. If you have any other examples, do share.

...lesser known in the Medical Device world...

Kevin, if you have experience in the Medical Device arena, do share. I am currently doing the same thing at my current employer (see Medical Device forum) and your inputs would be valuable.

For those interested, a variation on the ‘giddy girl’ test:

Fallacy Of 100% Inspection

Exercise:
º Count the letter “E” in the three paragraphs inside the frame
º Total and record the number
º You have 5 minutes to complete the exercise

SUPPLY OF OXYGEN AND REMOVAL OF CARBON DIOXIDE RARELY ARE CONSTRAINTS FOR AIR VOLUME; THE MOST COMMON CONSTRAINTS ARE ODOR REMOVAL OR TEMPERATURE CONTROL. THE AMERICAN SOCIETY OF HEATING, REFRIGERATION, AND AIR CONDITIONING ENGINEERS’ (ASHRAE) STANDARDS ARE GIVEN IN AIR VOLUME PER PERSON-MIN. TO ALLOW FOR VARYING OCCUPANCIES DURING A 24 HR PERIOD. AREAS WITH SMOKING REQUIRE MORE AIR CHANGES.

IT IS EXPENSIVE TO BRING SUPPLY AIR TO DESIRED TEMPERATURE, HUMIDITY, AND QUALITY; MOVE IT THROUGH A SPACE; AND THEN THROW IT AWAY. VENTILATION MAY USE AS MUCH AS 50% OF THE ENERGY REQUIREMENTS OF AN OFFICE BUILDING. REUSE THE AIR. THE RECYCLED AIR, PROCESSED TO REMOVE POLLUTANTS AND ODORS, THEN IS MIXED WITH OUTSIDE AIR (FORMALLY CALLED FRESH AIR) AND BROUGHT TO THE DESIRED VALUES OF TEMPERATURE AND HUMIDITY. FOR INFORMATION ON AIR QUALITY, SEE CHAPTER 6.13.

FILTERS AND PRECIPITATORS ARE USED TO REMOVE CONTAMINANTS AND ODORS. REMOVE CONTAMINANTS LOCALLY (SUCH AS THROUGHT EXHAUST HOODS) RATHER THAN LETTING THEM SPREAD AND THEN HAVING TO PROCESS MANY TIMES THE VOLUME OF AIR WITH GENERAL VENTILATION PROCEDURES. BE SURE THE WORKER’S BREATHING ZONE IS NOT POSITIONED BETWEEN THE FUME SOURCE AND THE HOOD. EXHAUST AIR FROM “CLEAN” AREAS (SUCH AS OFFICES) CAN BE USED WITHOUT PROCESSING AS INPUT FOR LESS CRITICAL AREAS (SUCH AS PAINT BOOTHS, STORAGE). WARM AIR, WHEN RUN THROUGH HEAT EXCHANGERS, CAN PREVENT INPUT AIR AND THUS REDUCE HEATING LOADS. HEATED AIR NORMALLY WILL BE TRAPPED (STRATIFIED) NEAR THE CEILING. IN WINTER, USE A HEAT INVERTER (FAN AT THE TOP OF A VERTICAL DUCT) TO BRING THIS WARM AIR DOWN TO THE LEVEL OF THE PEOPLE. IN SUMMER, LET THE HEATED AIR REMAIN IN A STAGNANT UPPER LAYER AND KEEP THE AIR CONDITIONING DOWN AT THE LEVEL OF THE PEOPLE.

Regards,
Don

What's the accuracy supposed to be? 80%?

Never mind - re-read and it's 85%. Excuse me!

But - since I'm EDITING, I'll also say this:

I wonder what the definition of QUALITY was back centuries ago when - well, take the Aztecs or folks 5,000 years ago (language is such a wonderful thing!). I wonder how they defined QUALITY or what their equivalent 'word' was. The concept had to come to exist at some point in time. Wonder when....

Any anthropology folks out there?

I wonder what the definition of QUALITY was back centuries ago when…

I am not certain a ‘definition’ existed, but ‘quality’ was, IMHO, considered.

Funny you should mention this. In some of the training sessions I give, I ask a similar question. It goes something like this:

Prior to modern manufacturing techniques, by what standard do you think quality was judged? After discussion, it usually boils down to: form, fit and function.

For example, when the transition from bronze to iron was occurring, it is safe to assume that both cultures co-existed for some period of time. A tribe had made the transition from bronze to iron, whilst his neighbors to the north (or south, either can be used as an example) were still trying to implement the new method (iron). Under the assumption that tribal warfare existed, during battle, the tribe with iron swords would have higher ‘quality’ weapons than the tribe with bronze. Not sharper, just better, using form, fit and function as the criteria. When the scales were leveled (both cultures made the transition to iron), the outcome depended more so upon other factors, such as ‘quality’ of commanders, ‘quality’ of manpower, ‘quality’ of tactics, etc. The same could be said when Europeans started trading goods with Native Americans. The tribes with the European steel weapons were superior in war to those tribes still using stone weapons. The steel knives were higher ‘quality’ because they better satisfied the form, fit and function criteria.

As artisans and craftsmen were used as a source of goods, the craftsman knew that if his goods did not meet the buyer’s expectations, his livelihood was threatened. Going back to the sword example, if a particular craftsman’s sword failed in battle (broken, too short, dull or whatever), he knew that one, he had lost a customer (dead) and two, if word spread that his swords were not dependable in battle, he would not be getting new customers. After all, who would use a sword of questionable ‘quality’ when his life literally depended on it. The same could be said for housing, furniture or food.

In a war based culture, the artisan that developed superior (higher ‘quality’) weapons, his reputation was hailed. When his work was inferior, he was beheaded, or some other sort of example setting. This particular aspect of form, fit and function exists even to this day (more bang for the buck is the common expression used). Case in point: Mutually Assured Destruction (MAD) policy at the heart of the US-USSR arms race. Therefore, the artisan took great care in assuring his product was superior to his competitor, through probably what we consider benchmarking, inspection and testing.

I know there are examples of ‘quality’ as it related to measurement and specifications. After all, the pyramids and such were not built as they were through serendipity. There had to have been measurements taken and if the measurements were not within a range (tolerance), they were redone. I do not have specific examples to share, just logic REQUIRES that it had to be.

Regards,
Don

For those interested, you can try this.

I had forgotten about this paper. It is an interesting read.

Regards,
Don
This is the paper that was at the now dead link.

With all due respect to Crosby, Deming, et.al. IMO the idea of zero defects or zero variation or any of these other "zero" philosophies is, in most practical situations, impossible to achieve and therefore doomed to fail to achieve its stated goal. Unless God is running your process, it ain't gonna happen. Humans are designing and running our processes and with near statistical certainty, given enough time/opprtunities, humans will make mistakes, period. A practical goal is reducing them as much as is both practical and possible in the given circumstance, and in doing so one should consider all of the tools in the Q toolbox, including old-fashioned inspection, and pick those that best fit the situation. My 2 pennies worth...

Yep, Zero Defects as a destination or a direction. Picking it as a destination, you won't be getting there. Picking it as a direction, well, that's more plausible. Zero Defects in the words of James Harrington is "...12 Sigma...".

As Don used to say, Zero Defects is a "noble concept", but doesn't really offer much else.

Kevin

Zero defects is just as clear to me as ,say, continuous or continual improvement. As for six sigma or better that only exists in the . . . well it doesn't exist.

Zero defects is just as clear to me as ,say, continuous or continual improvement.
I agree zero defects is, at this point in time, not technically realistic in most interpretive situations. If you do metal stampings, for example, with consideration to size and specific processes (i.e.: Stamping out a metal bumper for a car from not so long ago when bumpers were steel {was high volume} vs. stamping a blank circuit card {yes - material differences}) plays a big part. Application and all that. Point is, in large part this is a matter of definitions of what zero defects is with consideration to the process, its complexities as well as aspects such application and reliability requirements. Of course, VOLUME is always a BIG factor.

I've worked in companies where the product was low volume, high reliability. Mil Spec electronics where in a few cases we were talking in less than 100 units. But - they could NOT fail in use. Period. The potential for 'significant damage' (including the probability of multiple deaths) was too great. So - we delivered 100% tested, burned in (to kill 'infant mortality' issues) and triple verified. This was back in the late 1980's when everyone was complaining about US$500 hammers and such. There was so much testing required in every contract - extending to 'non-potential significant damage' items. So - In this respect we delivered Zero Defects product. Or so we thought. Field data over the years showed that, as happened with the last shuttle disaster, Zero Defects is a dream. Defectives do get out and 'bad' decisions are made.

There are so many factors involed that when talking about Zero Defects one has to look at the situation as a whole. In the automotive world, especially, the theory of Zero Defects typically is simply silly. From the aspect of volumes (i.e.: 1M a year?) to companies which accept contracts (poor APQP Contract Review {Ummm, I think we can do this vs. expressing valid concerns} - and I see this a lot in print reviews {sometimes as simple as whether or not a.) Tolerances NOT expressed, or b.) Can our equipment realistically hold the tolerances}). I've seen companies agree to GD&T prints which had no - I repeat NO - person in the engineering staff, or the layout lab staff, who really understood GD&T well enough to even know whether they could MEASURE the darn thing, much less whether they could manufacture it to the print.

Another basic factor is complexity and 'end point'. Are we talking resisters in the hundreds of millions (if we have a PPM of 5 we're doing OK) or are we talking a completed assembly?

My point is, when one starts talking Zero Defects, one has to take into consideration the context.

I have been involved in automotive for years and typically what I have seen is a requirement of Zero Defectives delivered whether piece part or assembly. This said, in this case we're talking about defectives delivered vs. defectives produced.

What is your definition of Zero Defects?

Your comments??

As for six sigma or better that only exists in the . . . well it doesn't exist.
I'm not sure what you're saying here.
Yep, Zero Defects as a destination or a direction. Picking it as a destination, you won't be getting there. Picking it as a direction, well, that's more plausible. Zero Defects in the words of James Harrington is "...12 Sigma...".

As Don used to say, Zero Defects is a "noble concept", but doesn't really offer much else.
A goal.

I agree zero defects is, at this point in time, not technically realistic in most interpretive situations.

If you look at it from the literal view point, I agree.

I've worked in companies where the product was low volume, high reliability. Mil Spec electronics where in a few cases we were talking in less than 100 units. But - they could NOT fail in use. Period. The potential for 'significant damage' (including the probability of multiple deaths) was too great. So - we delivered 100% tested, burned in (to kill 'infant mortality' issues) and triple verified. This was back in the late 1980's when everyone was complaining about US$500 hammers and such. There was so much testing required in every contract - extending to 'non-potential significant damage' items. So - In this respect we delivered Zero Defects product. Or so we thought. Field data over the years showed that, as happened with the last shuttle disaster, Zero Defects is a dream. Defectives do get out and 'bad' decisions are made.

Been there, Done that. Which ultimatley led to the expression,"test until it passes". Two firms I know of here in OK was convicted of falsifying test data. There is a downside to everything.

In the automotive world, especially, the theory of Zero Defects typically is simply silly.

Again if you interpret "Zero Defects" literally, I agree.


What is your definition of Zero Defects?

Your comments??

Zero defects is and always has been, a goal.


I'm not sure what you're saying here.

Six sigma does not exist in statistical text books nor is it taught in ststistical ciurses. Furthermore, Juran, Crosby and Deming did/do not condone this practice. It only exists in the minds of so-called process control gurus. IMO 6S is nothing more than shape shifting and masks the true appearance of the process.


A goal.

I Agree, Zero Defects is a goal.

It has taken me a while to formulate this, and I'm still not sure it will make sense (and, since six sigma has already been mentioned), here goes...

Ok, the Crosby approach has been around for a while, while six sigma, as a combination of mostly old tools, is relatively new.

One talks about confromance to requirements, the other measures defects as parts per million.

During the Southeast Quality Conference a few months ago, there was an excellent presentation by a six sigma "guru" (from Boeing, if memory serves). He spent a lot of his talk discussing the importance of defining all of the chances for errors to occur (the million that the parts are from, so to speak). To my mind, there are two ways of reaching the vaunted six sigma: either reduce errors, or find every nit-picking possible mistake to build your "million possibilities". So, comparing one six sigma system to another on the basis of sigma seems, to me at least, questionable.

On the other hand, Crosby talks about requirements, and a little about requirements that are not expressly presented (It may not say in the spec that the product has to arrive intact, but the customer will complain if it doesn't).

So, to me, both approaches have value, but "the truth", the optimal approach to outlining requirements, or the millions, lies somewhere between the two approaches, depending on the situation.



Craig

I wonder what the definition of QUALITY was back centuries ago when - well, take the Aztecs or folks 5,000 years ago (language is such a wonderful thing!). I wonder how they defined QUALITY or what their equivalent 'word' was. The concept had to come to exist at some point in time. Wonder when.... Any anthropology folks out there?

I believe that is where the "perception of customer satisfaction" measure came from. . . if your customer (The Grand Pu-Bah) wasn't satisfied, you had your head lopped off. This cut down on the poor performers in the supply chain, and started the continual improvement effort in a dramatic way.

Pardon the levity Marc. . .but I couldn't help myself. . . :topic:

During the Southeast Quality Conference a few months ago, there was an excellent presentation by a six sigma "guru" (from Boeing, if memory serves). He spent a lot of his talk discussing the importance of defining all of the chances for errors to occur (the million that the parts are from, so to speak). To my mind, there are two ways of reaching the vaunted six sigma: either reduce errors, or find every nit-picking possible mistake to build your "million possibilities". So, comparing one six sigma system to another on the basis of sigma seems, to me at least, questionable.

On the other hand, Crosby talks about requirements, and a little about requirements that are not expressly presented (It may not say in the spec that the product has to arrive intact, but the customer will complain if it doesn't).

So, to me, both approaches have value, but "the truth", the optimal approach to outlining requirements, or the millions, lies somewhere between the two approaches, depending on the situation.Craig

Craig,

Very observant comment about the opportunities part of the Six Sigma formula. The defect part of the formula is normally easy to define. I have yet to find an article on Six Sigma that defines how to count the opportunities.

I remember reading about Philip Crosby defending his zero defect concept. His belief was that you should strive for zero defects and not necessarily that you would ever get there. Marc talked about the shuttle's goal of zero defects. Another example of a zero defect goal is lost time injuries or, even worse, fatalities. As evidenced by the usual plant signs of days since last lost time accident, the goal is zero injuries. No one would ever set a goal of x nunber of lost time injuries per year much less set a goal for fatalities that was anything more than zero.

Besides the practical side of zero defects not being achievable in many cases, there is also the cost/benefit of even attempting unless it is a critical enough need.

Bill Pflanz

Bill, even further, what about the cost of defining the "millions" for six sigma?

NASA, it seems, did not consider the foam falling from the shuttle fuel tank a defect - it kept the fuel cool until it was needed, didn't it (assuming that was its function)? The foam had come off before during previous launches. Would the six sigma metric have encouraged placing this flaw in the "defect" pile? Would it encouage other things to be labeled as defects, when there really is little or no risk of assembly failure, or the cost is miniscule??


Bill, your mention of cost/benefit goes right to the heart of real-world application of these concepts. Economists will tell us that a business person will stop collecting data when the cost of that data exceeds the benefits. Sometimes, the cost of a particular defect happening is unknown. Heck, sometimes it is unclear if a particular state of being is a defect at all (this might be what happened to Marc's circuit boards?).

What, then, is the answer? Well, there is no perfect one. Experience and smarts are, in the end, our only defense. But, we have all of these great tools....

Craig

BTW, thanks, Bill, for the nice comments.