|
THE SEARCH FOR COMMON SENSE
Dear ASTE: Where have you been? I've been looking for a new newsletter from you for a long time. My company has lost its collective mind!! Entire departments have been created that never existed
before. And most of them exist to feed themselves, like some alien thing that sucks the life out of the company and all the employees within. They send out representatives that act as their tentacles and
they attach themselves to you and your environment like giant drinking straws draining off all common sense and possibilities of doing a good job. Common sense is no longer common!
Everything done goes into feeding these Gods of Statistics. I know statistics, if used appropriately, serves a purpose in life and in industry, but the way I see it being used is killing me and ultimately will
destroy our product. And anyone, and I mean anyone, who speaks out against the absolutely insane way it has taken over may as well fall heart first on a rat tail file. It is political suicide to challenge the
Gods of Stats. They calculate the measurement uncertainty of our inspection techniques, then instead of correcting them or even treating them individually, they just add it in as one lump number to the
final deviation from perfection of our product. This has the effect of causing perfectly good parts to be rejected. Since less of the yield of our product is accepted, and little effort is made to correct the
problems as individually treatable symptoms of the overall uncertainty, the situation continues on and adds to the overall cost of our product. We're loosing market share to our competitors, who have
yet to institute such a system. My only hope is that they catch up soon, so that they too can reject a bunch of perfectly functional parts. Help! Dear ASTE Member:
Read the next letter.
___________________________________________________________________ Dear
ASTE:
The statisticians have invaded my company and eliminated every maximum material condition symbol from every drawing we create. They say that there is no gage repeatability with the MMC
concept and that there is no way they can chart bonus tolerance on their Statistical Process Control charts. They don't care that functionality allows additional tolerance based on larger hole and smaller
shaft sizes. They don't care that these things allow reduced product costs that we can pass on to the customer. They don't care that the elimination of the MMC concept makes all those perfectly
functional fixed size gages and fixtures no longer allowable to use and instead asks us to use expanding fixtures for everything. They say this increases repeatability and reproducibility, but I've
used these things and I know that even if we could afford them (they're very expensive) these expanding fixturing and gaging elements aren't repeatable and the data they eventually give us is not
reproducible. Sometimes when you expand a fixturing or gaging element it moves the part and at other times it just locks up on the side of a feature not moving the part at all. And I'm not talking
about different parts either. You can use the same checking fixture on the same part six times and get six different measurement results. And since the mating parts in the assembly don't expand, we aren't
recreating the functional situation in any of these expanded locations and orientations, so what are we accomplishing? Also, since we can't use fixed size gages for anything, we are forced to use a Coordinate Measuring
Machine for everything, which is what they want so that they can again chart all the data, no matter how poorly and non-functionally it has been collected. Which brings me to my next problem. All the
CMM operators have been told by the CMM sellers that fixtures are entirely unnecessary when using a
CMM, to just probe a few random points on the surfaces. Forget all the studies that have
been done that say you need over a hundred points within each square inch of surface to even achieve a 75% certainty of form. They are told to believe they can probe the datum features any old
place they choose and get the same datum planes and axes that the fixtures would give if probed while resting on the highest points of contact. What in the world is going on? I feel like one of the few
people around here with anything above the ears besides mushy oatmeal. There are others, but they've been intimidated so badly, they've decided to just go along to get along. But this is wrong, wrong, wrong!
I'll tell you what I think. I think we are creating a monster that doesn't use any common sense whatsoever. It doesn't care about how the parts function. It doesn't care about the rising cost of our
product. It just cares about the act of performing statistical calculations. It only cares about the drooling satisfaction it receives from rows and rows of ultimately functionally meaningless numbers it
can churn out and put on a chart to show the admiring masses. But what can we do? Merely the act of voicing a concern, can end your career. You're labeled as not a team player. But what we don't
realize is we aren't the team. The statisticians have become the team. All we are is the equipment, no more important than the copy machine and far less important than those damnable SPC charts that they carry around like the Holy Grail.
The thing that really grinds my peaches is that these guys have skills, just no common sense. They think the means is the end, not a tool to improve the end result, which should be the improvement of
the overall company and its products. Higher costs don't improve the company standing, especially if those higher costs are based on non- functional processes created only to feed the Stat Monster.
This thing has gotten way out of hand. Is there any hope for us?
Dear ASTE Member: There is nothing wrong with Statistics, as a discipline of mathematics, but there is often something
wrong with the way it is applied. And, of course, any suggestions for improvement made from the data collected is only as good, and as reliable as the data itself. You may find some comfort in the
fact that I have been reading a number of authors recently who say that they think foolproofing will someday make the need for many statistical procedures unnecessary. One example they give is using
fixtures which both inspect and hold a workpiece for the next manufacturing operation simultaneously. I love telling this story because it makes all those who have made statistics their God
nervous and consequently angry (no matter how badly or ineptly the statistics are practiced). They turn all red in the face and hiss through clenched teeth that this type of thing will never happen, and
anyone who would spread such a tale must not be a team player. Still, it is worth considering. If anything that can reduce costs, improve quality, save time and improve the mental health of a company's employees, it has to be worth a look.
What if all manufacturing and design considerations had to consider functionality first? Products would be designed to work great and be easy to manufacture. We wouldn't deny the manufacturing
people any of the functionally acceptable tolerances just so an SPC chart could turn out an acceptable statistical number. Yes, we would still monitor our processes, but not throw out parts that
don't fall within a functionally meaningless statistical range. There would be jobs for the statisticians to perform, but they would not be able to change fully functional, low cost, common sense
manufacturing procedures. All measurements could be directly related to whether or not a product would work well. Manufacturing and measurement procedures would be improved by targeting
specifics and changing them to be better, not just identifying a vague non- repeatability or uncertainty and adding that lump number to the measurement deviations of the product features. Product yields
would be greater, and yet all increases in yield would be of products that functioned just as well as those rejected by previous non-functional requirements. Fixturing would be used whenever possible
to recreate the way a product fits into its overall assembly. We would no longer take the word of nonsensical statistical data that claims to be able to interpolate points on a parts surface that have not
been probed or inspected. Gages would play a greater role in part inspection and, in many instances, the fixtures would become the gages. Variables data collectors like the coordinate measuring
machine would be used optimally, not alone, but with help, the help of fixturing devices and augmenting gages that would allow the datums to be formed in a more efficient manner than if the
CMM probes were brought in contact with a few surface points and from those randomly selected points datum planes and axes formed, pretending to be the same as, or as good as the high point
planes and the axes created by probing the high points.
Isn't it interesting how we will try to repair a manufacturing procedure based on data that is collected
in a grossly non-functional manner. And in doing so data collected after the ``correction" is often worse than the data collected before, or the data seems to be better, but the parts don't work as well
as before the ``correction". A resurgence in the use of and interest shown for gages and fixtures around the world in the last couple of years is probably not just a coincidence. I think some, many
perhaps, have just plain lost faith in these non-functional procedures, that don't seem to have a basis in good sound common sense. And some folks are running to procedures and tools that they know,
from their many years of solid experience and judgment, are trustworthy. Most often they don't want to get rid of the new procedures all together. They know they have some merit. But what they are
rebelling against is a mindless adherence to this new cult-like subservience to procedures and data that seems to have no basis in good sound logic. They want to test it, to examine it, with some of the
tried and tangible tools and techniques they grew up with. And when they do that and the information comes out different from the information they are being fed by these techniques that have been
forced down their throats, they want to discuss it, improve it, and not just blindly go where they are being forced to go no matter what the consequences just because it happens to be the program
of-the-day. These programs come and go, but years of experience and logic and good common sense are things we have come to rely on and need to use. Search for the sense in an act, don't just
perform it. Use all your judgment and reason to determine if something has merit before accepting it as your mission in life. Remember, you are a sentient being, the culmination of a lifetime's knowledge.
Don't just throw that away. ___________________________________________________________________ Dear ASTE:
I'm deep in high cotton and happier than a pig in slop. I just got a contract to teach a thousand
people Geometric Dimensioning and Tolerancing. And it's so close to my homestead I can mosey on home at noon for biscuits and red eye gravy. Can you loan me a video on this subject? I been
cogitatin on this and since I never done geometric anything before, I figure I ought to study up on it before I commence to doin it.
Morty, New York City Dear Morty: You're in something deeper than high cotton. ___________________________________________________________________
DUDES: Like what is GD&T? It sounds pretty cosmic. Does it have to do with channeling? I know this chick who does a transformation into a three thousand year old warrior from Atlantis who has lived on in
the winds of time and space. He's called Gaga Din Tin or GD&T for short. Dude, Mississippi Dude! Do they let you talk like that in Mississippi? |
