iSixSigma

Control charting

Six Sigma – iSixSigma Forums Old Forums General Control charting

Viewing 58 posts - 1 through 58 (of 58 total)
  • Author
    Posts
  • #49085

    Singh
    Participant

    Hello,
                May you give some examples whereby presence of special cause in a process is beneficial and these are required to be integrated in to the process?
    Regards.
    Kanwal

    0
    #167130

    Dr. Scott
    Participant

    Kanwal,
    Yes I can. If you give me your email address or email me at [email protected], then I will be pleased to show you a real example of how special causes can be good.
    You must remember special causes are simply as the name suggests. They are special or unusual. They are not always bad as many here think. You can use what you learn from bad special causes to eliminate the bad root cause. But you must also use what you learn from good special causes to standardize the root cause.
    Email me and I will send you the real data that might help what I explained above.
    Regards,
    Dr. Scott

    0
    #167148

    Craig
    Participant

    I can’t say that I have heard of “good” special causes, but I think as long as it is predictable, you can react to it accordingly.
    Take an etch process for example that shows a linear change in etch rate as you process more and more material through it. You can predict the etch rate and adjust the time for the next batch accordingy.
    I would much rather have a process that simply stays on target, where no adjustment is needed. If the special cause is predictable, I suppose you could rank it as “tolerable”.
    HACL

    0
    #167159

    Six Sigma Shooter
    Member

    I have a problem with your teminology.  If it is a predictable source of variation, it is a common cause, not special.  The failure to remove special causes from the process makes them common to the process.
    Secondly, there are so-called “good” special causes.  An unexplained improvement or reduction of variation or defects.  This is a good thing.  One would want to find out why it happened in order to make it happen all the time and predictable, thus making it a part of the common process.

    0
    #167163

    Craig
    Participant

    Common cause is associated with noise. (random variation). Special cause is associated with signal and  also is classified as assignable cause.
    An etch rate that degrades over time is assignable, or special cause. The slope of the line could be called the signal, and the variability around the line (residuals) would represent noise.
    Your statement about failing to remove special causes is somewhat confusing. Variation is either random or non-random. If you just sit back and watch non-random patterns of variation, that does not make them random. They are still non-random, special-cause sources of variation.
    HACL
     

    0
    #167170

    Terry
    Member

    I disagree …
    Special causes are the same as assignable causes of variation. In other words, they are predictable and not random. Common causes of variation are random, which is why they are common to any number of processes.

    0
    #167171

    Six Sigma Shooter
    Member

    Special causes are those that are not common to the process.  When they happen, they are a “surprise.” They are called “assignable,” because it is usually easier to identify and assign a cause for the variation when proper investigation is conducted.  Their very nature makes the process unpredictable.  Common causes are those that are common to the process when a process is stable and in control, though their source may be unknown.  One of the great values of a control chart is that when a process is in control and capable, displaying only common causes of variation, it is a predicatable process that it will operate within its control limits.  Do not confuse the random nature of common cause variation within the control limits of a process that is in control and capable with special cause variation that enters into a process from outside of its normal operation.

    0
    #167173

    Terry
    Member

    The idea that the cause of an instability is harder to remove than a common cause variation must come from working in an ivory tower. This is not directed at you, but to the author of this ‘assingable implies easy to find nonesense.’ Instability is notoriously difficult to rectify because it can be the result of non-linearity.
    On the other hand common causes of variation can be reduced easiliy, simply by moving the process to a flat part of the response surface.
    The definition you use also raises the question of what you call an assignable cause of variation – an instability-  before you’ve found the underlying cause? What do you call it in the meantime and what do you write on the control chart  – an ‘as yet unassignable cause of variation?’
    Thank you for the exchange, I’ll leave the last word to you!

    0
    #167181

    Six Sigma Shooter
    Member

    First, I said that special causes are often easier to identify, which they usually are, in my experience.  Shewhart also thought so when he developed the control chart based upon extensive study.
    Second, Dr. Deming was said to be many things, but living in an ivory tower is not one of them.  Actually, he spent a lot of time in his basement office when he was at his home in Washington DC. ;-)
    Saying that removing common cause variation is easy, sounds good in theory.  Not so easy in practice.  Especially when you have inherent variability within very expensive equipment that would have to be upgraded or replaced in order to reduce the common cause variation.  And try running a DOE on one of them that is operating 24/7 in high volume operations.
    What do you call a special cause while under investigation? I could care less, as long as you do the investigation, find the root cause and remove it,  What I write on the control chart is “under investigation” and I leave it there until I identify the root cause and remove it.  If it can’t be removed, it stays and becomes a part of the process.  Now, there are times when “freak outliers” occur, but they are rare.  Only seen a couple in my experience (one being in the red beads experiment – a very stable and in control process).
    In the end, it’s interesting to engage in these debates about theory and practice.  The important thing is to do the work, improve the process and reduce variation.  All else is an exercise in mental jousting.  Our job is to first identify and get rid of the special causes of variation so that we have a stable and in control process.  Then, we either work to reduce the common causes of variation, or we live with it – generally a management decision based upon economics.

    0
    #167183

    karl
    Participant

    Six Sigma Shooter:”Saying that removing common cause variation is easy, sounds good in theory. Not so easy in practice.””Now, there are times when “freak outliers” occur, but they are rare. Only seen a couple in my experience (one being in the red beads experiment – a very stable and in control process).”What other process do you have experience with – the catapault? :-)Karl

    0
    #167184

    Six Sigma Shooter
    Member

    Karl,
    I’d  be equally justified in asking you the same question.  If you have a point other than trying to denegrate ones experience by innuendo, please post it.  Otherwise, your post has no inherent value.  Do some basic investigation and you can find my experience level and background.
     

    0
    #167190

    Lou Smidt
    Participant

    Hilarious Karl – you can now go and search for erudite articles by Six Sigma Shooter!I wonder what the special cause of variation was in the red bead experiment – oops! dropped the beads again!

    0
    #167195

    Shooter
    Member

    Lou,
    If you understand what a freak outlier is, you’d know that it is common cause variation, not special cause.  Deming was very keen on making sure of it being an in control and stable process, which is a key factor in the red beads experiment.  Had they been dropped, Deming would have made them do another pull. 
    It was at his seminar in 1992 where the instance I referenced occurred – one of only four occurrences of freak outliers that he’d experienced in his many years of red bead demonstrations.  He kept very good records of all of the pulls over the years and he noted that there was a significant difference in the number of red beads pulled when he used different paddles, though he meticulously maintained the same number of red and white beads in the bowl.
    Good luck in your endevours,
    Shooter

    0
    #167197

    Mikel
    Member

    As Deming pointed out, the workers should be able to correct special causes and the remaining 96% of problems are common causes and should be addressed by management.

    0
    #167199

    Six Sigma Shooter
    Member

    Exactly, Stan.  And I think this goes directly to the point regarding the ease of removing special versus common causes of variation.  Trying to get management to take ownership and to get out of their red beads mentality – well, all I can say is, lots of luck, especially in western cultures.  To me, anyone saying that special causes are predictable (beyond the generalization that they will occur) and that common cause variation is “easy to remove,” either has little experience, little understanding or both.  Oops, there I go, being “erudite” again.  Too funny!
    Shooter

    0
    #167200

    Taylor
    Participant

    My mentality has always been that process variation whether Special, or Common is an opportunity. Real Innovation to process improvement begins with root cause analysis of these types. It makes no difference if they are bad or not, something happened, and it must be understood in order to repeat it or not repeat it. The very fact variation occured has probability it will occur again.
    I often wonder where some of these guys work, because management must be really frustrated with there hands off approach to problem solving.
     

    0
    #167202

    Six Sigma Shooter
    Member

    AMEN!  As Deming often said, management’s job is the continual improvement of the system and the reduction of process variation.  Understand the sources of variation and gain profound knowledge. If it’s a negative, remove it.  If it’s a positive, incorporate it.  Either way, investigate and understand it and its impacts on the system. 
    I’m really not sure that their management is frustrated with them,, or they would resolve the issue.  Management by abdication of management’s role is still alive and well.
    We’re getting too caught up in the mental gymnastics and forgetting the “why we’re doing it.”  Leave the cozy cubicle and the cool software behind, get out into the workplace, work with the people in the process to really help them in a colaborative manner, and watch them do great things.  Unfortunately, many of the current breed I’ve come across are more interested in mandating change after telling everyone “I’m here to help,” making sure they get the prerequisite “savings” and then moving on to create havoc in their latest claim to glory project. ”
    I had one Black Belt candidate tell me he had a very good team in that they did everything he told them to do.  This, after he’d completed a cause and effect on their process for them, by himself, and then went out to the factory floor, flashed it in front of them on an individual basis while they were working on assembling cars, and asked if they agreed with it. We had a very long talk about his style of teamwork.

    0
    #167204

    Deanb
    Participant

    Shooter,I have seen far too much of that “glory grabbing” too. It is ugly, and damaging.Dean

    0
    #167213

    Lou Smidt
    Participant

    Shooter,What does ‘again’ mean to you in my last post. Do you regard two freak outliers in a row as a common cause?I apologise for my humour, I just thought it was funny you seemed to regard Deming’s red bead experiment as an example of a real process.Taking Stan’s point, just because there are many more common causes of variation, it does necessarily follow that they are technically more difficult to reduce.On the other hand, in my own experience, I’ve come across many examples of instabilities that were very difficult to eliminate, especially when they present with different symptoms each time.I also find it somewhat ironic you consider special (assignable) causes easy to identify when much of six sigma is devoted to a ‘shift allowance’ of as much as 1.5 sigma, presumably on the basis that these sources of variation can never be resolved.Good luck to you!Lou

    0
    #167218

    Six Sigma Shooter
    Member

    Nice try, Lou, but your response is nonesensical weaving and dodging.  I’ll let your post stand for what it is.  Taking it on point by point is . . . pointless.  It can easily be seen for what it is.  Another product of bad training, bad learning, or both.

    0
    #167219

    Deanb
    Participant

    Shooter,My previous post did not get posted, so here it is again.The behavior you described:-saying they are there to help, (implying shared input and decisions) then mandating decisions.This is a form of false pretense. In a study I was a part of a few yrs ago, this was one of the most destructive managerial behaviors observed. It correlated with worker dissatisfaction and with reductions in open communication, process capability, and ability to find root causes. This was definitely one of the king kongs of anti-quality management behavior!This is an error caused solely by ignorance, and it is just as important as a technical error, if not more so, because technical errors can be corrected more easily. I believe that healthy alternatives to this behavior need to be incorporated into SS BOKs and taught as a high priority in BB certification.Dean

    0
    #167220

    Fake ATI Alert
    Participant

    Who  cares about the SS-BOKs?
    Is  there  an  organization (standardized  reference) like ASQ,villanova,etc (similar to  the ISO Organization )?
    thanks  and  regards
     

    0
    #167221

    Six Sigma Shooter
    Member

    Dean,
    You’re spot on.  Even with the best of positive intents, some pretty destructive things take place.  When you couple this type of “help” with an elite group of people, set apart from the rest of the organization as a task force or special team, with special titles and their own cozy office, it just makes for a very bad mix.  At best, it leads to a false start and at worst, it completely destroys the continual improvement  initiative. 
    If management would actually put themselves in the shoes of the people on the recieving end of the systems they develop, I think many of these problems would disappear.  Just a good dose of some good old fashioned empathy and human psychology would go a long way. 
    Shooter

    0
    #167222

    Deanb
    Participant

    I have often said that the experience of doing a BB project as part of certification is not as relevant as having a BB project done to them.

    0
    #167223

    Real Data?
    Participant

    Gee Dr. Scott, could you show all of us some real data?I can’t imagine that when a process shows a special cause that is to
    our benefit, that we can actually learn from it.

    0
    #167225

    Six Sigma Shooter
    Member

    I find that organizational BOKs tend to be self limiting and prevent a true holistic approach.  They have a narrow view, which only encompasses their “world.”
     
    The complexities of continual improvement involve many different disciplines and skill sets, or BOKs, including:  psychology, project management, benchmarking, strategic analysis, critical thinking . . . the list goes on.  Unfortunately, six sigma, as it has evolved, has limited its scope to the hard skills of data gathering and analysis, and has pretty much left the softer skills in the fog.  With it’s focus on the statistical hard skills, it is very limited.  And when you look at many of the failed deployments – be it TQM, Quality Circles, Lean, whatever – it’s not due to a lack of hard skills knowledge, but due to the lack of the soft skills knowledge and application.
     
    I once brought up the issue that six sigma training should also include, at a bare minimum, project management, group dynamics and team facilitation education.  The answer I got back was, “Everyone knows how to do all that sort of stuff, don’t they?”  This assumption couldn’t be more wrong.
     
    If we truly want to achieve a state of continual improvement, we must seek continual education and learning, adapting what we learn to our reality, and apply what we’ve learned and adapted to the continual improvement of our business.  It must be a holistic approach, in my opinion.
     
    Shooter

    0
    #167226

    Taylor
    Participant

    RD
    Your Kidding, Right?

    0
    #167228

    Six Sigma Shooter
    Member

    I’m not Dr. Scott, but I can give you what I think to be a good example.
    Working with a process, we discovered that there was a special cause on the low end of a p-chart: less defects.  It just showed up on the night shift data one day.  Digging into it, we found that one of the workers wasn’t following the established process SOPs.  The process was part of a PCB masking process.  The SOPs called for the cleaning of the screen used in the masking process, first by wiping it down with a sterile rag, and then blowing the screen off with plant supplied air.  The worker had stopped using the air to blow off dust particles on the screen left by the rag. 
    Come to find out, the facilities guys were injecting oil into the plant air system for air driven tools used in their machine assembly processes (a very common practice with air driven tools to keep the operators from having to oil their tools) The result was a fine mist of oil was being injected onto the masking screen, causing solder bridges and other defects due to poor masking.  This issue had plagued the client for years and no one could figure it out until this special cause cropped up and they were able to focus in on its root cause.
     

    0
    #167229

    Six Sigma Shooter
    Member

    Chad,
    That was my reaction, but I figured, what the heck, I’ll bite.
    Regards,
    Shooter

    0
    #167231

    Deanb
    Participant

    Shooter. Excellent example. Now, if your process had achieved zero special cause variation (and your worker used contaminated air to clean screens), doesn’t it follow that there would actually be MORE defects in this process?The same argument is made in transactional processes, such as marketing, that some intentional variation is necessary to “discover” opportunities. Especially in transactional areas, I have found that certain variations can have different economic meanings.The notion that all variation is bad and must be eliminated perhaps needs qualified and tweaked a bit, to instead say that all variation first needs to be understood before acted upon.

    0
    #167233

    Fake ATI Alert
    Participant

    Excellent explanation,
    thank  you

    0
    #167236

    Six Sigma Shooter
    Member

    Thanks, Dean.  The really interesting thing was that the worker had circumvented the SOPs in the interest of schedule pressures.  He felt it that he would get more stuff out the door during his shift, albeit with the given poor quality, than if he followed the process SOPs.  Who knew?  And the thing that he feared most – that his lack of adherence to SOPs would get him in trouble – actually ended up being a big win for himself, his coworkers and the company, a manufacturer of office machines and a two time winner of the Deming Prize (this example had nothing to do with their winning).  The defect rate due to the masking process dropped to almost nothing, and also allowed them to zero in on the other sources of PCB and solder defects.
     
    Shooter

    0
    #167239

    Heebeegeebee BB
    Participant

    I think “Real data? ” fell on his head…

    0
    #167240

    Six Sigma Shooter
    Member

    Not sure that “Real Data” was for real in asking the question.  I think it may have been an effort to bait Dr. Scott.

    0
    #167245

    Deanb
    Participant

    Another needed revision to SS core theory is:”SS is not really a data driven hard science, and it fails science if it pretends to be. Instead it is a rational, evidence based and people driven social science.”Data and statistics are merely forms of evidence, among many, that are employed by people.

    0
    #167246

    Paulo
    Participant

    Let me try to understand the process. The operator on the night shift used the sterile rag to wipe the mask down, but did not blow off the lint and this led to less bridging?Is this a good way to make less defects, to put more lint on the mask? I hope you don’t mind me asking because we use lint free cloths. Perhaps this is a big mistake on our part.By the way, we don’t put oil in our air lines, only air.Thank you in advance.Paulo

    0
    #167248

    Six Sigma Shooter
    Member

    Paulo,
    Obviously, lint is not a desirable thing in the process.  They continued to use the sterile rags.  They set up their own air system (separate compressor, air lines, traps and filters), without oil added, and they made sure the moisture traps were rigorously maintained to prevent moisture from getting into the line.  They also cut down the PSI on the line from the typical plant 90 psi to 45 psi.  Even though their rags were sterile and supposed to be lint free, they found that there was still some that got into the process, along with other airborne particulates, since they were not in a clean room environment.
    Shooter

    0
    #167249

    Heebeegeebee BB
    Participant

    Hey Paulo, Long time no see!
    Last I heard fromyou, you were fighting with Nikki over a bag of diamonds and she had Arzt’s spider collection…
    Man, Hurley covered that up.

    0
    #167251

    Taylor
    Participant

    Paulo, the point to the story was to point out that just because variation happens in a process it is not always bad. In this process an operator stopped following SOP and suddenly his product had higher pass/fail ratio, this prompted an “Opportunity” of discovery which led the team to discover the original problem; oil in the air. This is a very easy fix long term, but it took a sudden “Special Cause” to shed light on the original problem.

    0
    #167259

    Dr. Scott
    Participant

    Super Waste1 1663.632 1665.423 1684.724 1646.965 1522.476 1601.937 1580.198 1591.221 1484.232 1677.363 1905.384 1780.845 1466.706 1745.527 1731.998 1666.031 1644.052 1749.823 1458.334 1568.925 1808.506 1642.087 1709.958 1850.211 1519.422 1638.053 1702.324 1756.995 1569.646 1660.157 1558.848 1743.651 1567.352 1681.393 1805.484 1603.475 1361.066 1879.987 1838.078 1574.051 1687.442 1515.633 1804.524 1494.525 1591.356 1596.647 1869.408 1847.131 1565.682 1591.233 1635.034 1584.775 1228.466 1824.527 1820.368 1612.191 1641.382 1683.293 1733.924 1746.885 1383.576 1755.087 1791.028 1557.891 1886.882 1698.663 1603.514 1465.495 1665.306 1655.487 1587.288 1619.931 1628.852 1739.303 1727.344 1465.865 1470.266 1767.217 1746.418 1671.631 1576.362 1780.713 1667.384 1807.305 1647.046 1801.117 1738.988 1618.041 1800.662 1689.473 1638.334 1722.535 1231.416 1830.977 1744.998 1592.071 1823.222 1845.333 1482.914 1704.205 1065.836 1787.407 1838.618 1653.981 1644.142 1661.893 1669.984 1538.265 1608.306 1764.887 1670.018 1698.131 1619.572 1552.973 1772.604 1648.725 1506.336 1701.197 1751.678 1615.711 1629.892 1779.563 1658.524 1787.885 1576.806 1518.717 1617.148 1888.221 1587.112 1580.653 1712.504 1553.705 1857.726 1896.987 1815.938 1512.371 1561.342 1473.393 1616.854 1940.195 1375.056 1583.607 1751.268 1499.111 1834.912 1749.893 1758.884 1501.285 1455.816 1873.407 1605.958 1615.441 1721.402 1591.573 1738.894 1678.135 1779.656 1641.257 1529.758 1747.071 1650.682 1816.353 1526.104 1499.845 1175.456 1550.127 1503.078 1642.321 1687.172 1704.013 1859.524 1620.915 1449.896 1785.967 1681.268 1844.761 1496.662 1669.683 1822.924 1450.795 1415.836 1476.357 1750.318 1772.161 1762.832 1670.343 1607.264 1624.165 1607.386 1776.907 1629.368 1921.351 1590.122 1996.023 1719.214 1856.075 1845.156 1638.367 1625.098 1750.821 1765.942 1628.873 1588.174 1990.125 1646.416 1618.637 1553.678 1582.231 1878.512 1776.713 1657.404 1644.335 1429.066 1660.577 1638.118 1678.771 1672.592 1886.443 1490.644 1512.355 1347.946 1603.787 1707.208 1778.781 1779.862 1663.913 1534.734 1622.215 1962.146 1908.127 1824.988 1729.741 1672.972 1852.083 1877.804 1631.165 1453.766 1708.297 1882.648 1708.241 1587.332 1639.193 1667.704 1589.285 1643.286 1646.087 1608.508 1635.261 1626.722 1884.413 1927.804 1704.975 1402.446 1648.297 1685.718 1513.731 1821.002 1736.183 1751.784 1855.345 1671.076 1959.017 1625.928 1770.881 1632.552 1819.523 1815.944 1739.345 1558.526 1669.497 1702.238 1577.381 1657.402 1809.013 1655.784 1689.945 1492.356 1754.397 1662.388 1922.501 1612.062 1943.523 1629.024 1815.615 1376.986 1554.757 1583.448 1816.351 1551.362 1754.813 1417.494 1660.035 1608.166 1658.047 1776.758 1649.721 1759.722 1641.983 1657.464 1822.745 1557.116 1513.527 1690.148 1657.391 1838.792 1619.263 1780.264 1528.495 1425.866 1721.187 1791.358 1913.791 1807.542 1693.123 1588.584 1692.675 1378.316 1666.927 1630.048 1728.821 1688.042 1711.073 1820.664 1612.355 1391.906 1658.087 1700.508 1416.961 1847.842 1756.703 1793.034 1892.345 1539.396 1812.757 1683.448 1755.861 1531.392 1695.603 1549.784 1544.735 1716.346 1773.937 1797.908 1523.901 1781.812 1597.033 1746.964 1898.215 1634.316 1733.117 1794.098 1398.151 1606.142 1679.003 1820.264 1879.355 1360.886 1610.207 1669.038 1887.511 1589.592 1704.743 1681.924 1656.935 1556.366 1651.187 1874.208 1803.551 1536.252 1652.403 1658.324 1739.495 1474.136 1482.157 1776.668 1767.161 1692.312 1554.753 1553.114 1600.225 1222.896 1615.617 1812.408 1794.781 1900.542 1445.303 1953.214 1916.485 1229.866 1729.407 1476.058 1768.881 1756.252 1762.313 1706.864 1771.555 1455.036 1873.947 1676.788 1426.801 1822.982 1702.873 1869.244 1653.255 1585.746 1656.147 1702.748 1527.041 1645.842 1936.513 1683.364 1504.775 1666.476 1700.867 1743.648 1771.091 1585.772 1706.493 1702.964 1712.675 1436.436 1711.067 1676.088 1693.131 1768.002 1653.903 1486.464 1642.105 1480.046 1762.497 1751.828 1600.371 1483.192 1795.553 1657.934 1643.095 1662.016 1537.177 1707.188 1616.971 1785.292 1942.063 1704.194 1897.765 1560.176 1559.127 1800.858 1779.371 1761.552 1680.963 1845.414 1553.165 1459.636 1498.817 1832.858 1679.781 1584.542 1802.353 1737.744 1711.155 1372.436 1804.477 1564.778 1630.051 1745.772 1755.083 1534.704 1762.925 1587.866 1742.507 1655.678 1914.321 1725.682 1703.813 1612.244 1743.245 1656.516 1726.597 1635.648 1713.651 1704.232 1505.533 1826.234 1732.855 1537.026 1797.997 1567.298 1527.451 1629.472 1683.223 1550.824 1743.655 1329.086 1877.647 1721.308 1764.071 1721.122 1688.443 1712.534 1635.465 1717.606 1816.137 1798.758 1743.051 1690.292 1784.933 1683.954 1588.265 1658.956 1656.037 1807.888 1657.421 1765.142 1754.463 1464.314 1698.305 1552.306 1631.157 1775.318 1477.101 1689.482 1678.583 1589.934 1566.175 1535.166 1803.537 1803.498 1788.671 1648.432 1796.423 1718.154 1764.115 1566.656 1922.837 1725.628 1610.621 1666.742 1456.923 1890.724 1398.535 1382.286 1731.127 1595.028 1579.431 1598.072 1495.113 1532.884 1748.635 1530.576 1776.117 1776.748 1601.931 1873.932 1726.273 1593.924 1623.675 1432.946 1632.267 1841.108 1562.341 1915.582 1744.823 1685.044 1927.705 1614.036 1866.547 1731.538 1634.681 1524.282 1489.223 1736.034 1820.175 1326.396 1671.787 1756.668 1716.841 1621.292 1693.933 1577.994 1669.565 1347.626 1810.357 1417.218 1665.561 1542.012 1577.753 1753.734 1775.245 1491.086 1813.717 1723.968 1752.861 1721.652 1462.963 1547.174 1618.675 1578.066 1971.427 1700.798 1716.361 1774.072 1833.283 1759.764 1859.285 1398.186 1544.137 2002.238 1484.311 1649.972 1716.003 1654.934 1531.975 1278.186 1619.077 1515.368 1658.361 1758.162 1862.133 1697.014 1748.385 1624.266 1731.317 1736.908 1835.201 1607.162 1596.383 1760.494 1921.285 1486.946 1519.197 1887.198 1689.061 1632.082 1583.453 1757.724 1798.835 1680.426 1730.347 1653.468 1669.941 1705.262 1861.353 1776.604 1736.325 1343.756 1559.417 1594.618 1852.771 1530.372 1877.363 1678.064 1902.815 1414.266 1634.507 1667.428 1971.251 1810.512 1764.723 1742.014 1555.865 1499.776 1631.337 1690.958 1729.831 1580.892 1621.433 1867.624 1650.655 1551.096 1811.507 1779.008 1694.691 1786.292 1642.993 1591.994 1538.635 1489.026 1676.927 1683.348 1597.531 1729.462 1826.113 1577.794 1479.375 1774.686 1484.997 1573.048 1781.971 1758.482 1603.673 1633.074 1785.355 1485.606 1702.667 1894.528 1711.851 1699.122 1818.033 1655.324 1682.125 1549.106 1957.247 1535.618 1611.761 1643.112 1601.493 1677.044 1587.765 1506.316 1749.847 1814.318 1870.681 1584.022 1792.003 1726.044 1613.105 1692.486 1667.577 1793.558 1894.811 1713.562 1733.963 1754.264 1716.105 1365.096 1673.987 1791.098 1569.731 2116.912 1863.793 1639.574 1820.275 1660.916 1739.647 1715.098 1654.561 1911.042 1601.853 1787.554 1836.415 1402.206 1680.147 1681.078 1627.541 1741.752 1789.673 1619.514 1708.755 1444.836 1627.327 1657.188 1814.461 1492.492 1725.343 1898.704 1578.025 1417.736 1957.147 1843.558 1875.941 1772.322 1922.723 1583.484 1630.055 1560.826 1561.897 1644.148 1643.351 1598.682 1665.693 1503.144 1798.745 1641.396 1608.737 1583.388 1614.921 1779.212 1747.483 1791.594 1834.235 1536.796 1672.957 1715.378 1777.621 1774.462 1810.853 1571.994 1797.455 1645.156 1825.097 1683.738 1732.991 1683.672 1588.333 1740.914 1755.835 1755.136 1679.477 1928.568 1741.63

    0
    #167260

    Dr. Scott
    Participant

    OK RD,
    There is your data. The first row is the shift supervisor; the second row is the average startup waste per shift. So you please do the analysis and answer the question posted.
    And I am sure it would be a treat for all of us if you could identity the source of the good special causes. And even a greater treat if you knew why they were good special causes (as I discovered).
    If you don not respond, then am I certain you are not worthy to. And please refrain from ever being disrespectful to me in the future.
    Regards,
    Dr. Scott

    0
    #167261

    Dr. Scott
    Participant

    Deanb,
    You are correct. And it is taught in my training and is one of the requirements of my certification.
    Your advice here and that of Shooter is good advice.
    Thanks,
    Dr. Scott

    0
    #167262

    Sorry Dad
    Member

    Gee Dad, can I grow up soon?
     

    0
    #167264

    Dr. Scott
    Participant

    That was very well said Shooter. I am impressed! And I have followed that approach in training all of my BBs and MBBs. Good to see that there are others out there that understand our work depends mostly on people, not just with software.
    Sincerely,
    Dr. Scott

    0
    #167265

    Deanb
    Participant

    Thank You Dr. Scott.

    0
    #167266

    Dr. Scott
    Participant

    To answer your question; seems to me you have a lot of growing to do. But if your have children, as I do, then I am sure you understand that your attitude will not help in your attempt to grow.
    I am still waiting for the results of your analysis.
    Regards,
    Dr. Scott

    0
    #167268

    Mikel
    Member

    “It must be a holistic approach, in my opinion.”
    Exactly !  What Deming referred to as “systems thinking”.  Quite a contrast to the project by project focus of Six Sigma. 
     

    0
    #167271

    Paulo
    Participant

    OK – I did not explain myself well.Even though you still had lint on the mask you were able to achieve very low defect counts. I think you mentioned zero. This is what I want on our process so it is tolerant to even more lint. (I made the mistake here in my last post.)When I showed our engineering manager your post, he also said lint will give more bridging and I told him I want a process that can tolerate much lint like in your experience.When I told him about the oil he said oil on the mask will make holes not bridging so I am a little confused. It is not a problem though because I still have to learn Six Sigma but I’m really glad I don’t have to take the trouble to learning engineering to make understand the process and to improve the process using the statics. Thank you very much for this enlightenment.Paulo

    0
    #167272

    Paulo
    Participant

    Eu sei que o homem?

    0
    #167273

    Craig
    Participant

    Kanwal,
    I have had processes where the variability suddenly increased, and of course this is evidence of the “bad” type of assignable cause. In one case, the spikes were merely due to intermittent production (stopping and starting of production due to lack of work in the area).  This example is from the wafer thinning process in the semiconductor industry. If I look at the low variability time intervals as the “good” assignable cause periods, my answer to your question is that we had to simulate a steady state operating mode during the intermittent production intervals. Generally, if the machine was idle for longer than X minutes, perform a warm-up procedure. Very simple, yet effective!
    A similar example comes from my SMT experience. Patterns of increased variation in solder paste volume were also from intermittent production. The worst thing you can do to a screen print process is to halt it for an extended period where the solder paste is allowed to sit and dry out. (The wait time in this case was from manual operator involvement for cleaning the bottom of the stencil). We had a high volume environment and there were no issues with lack of work in the area. For this particular process, we simply automated the stencil wiping. Again, I can say that the good assignable cause was running in a steady state mode, so we enabled this by implementing a quicker and more effecting cleaning process. It did cost a little extra for the rolls of lint free cloth on the machine, and the benefits easily outweighed this cost. Less variation in solder paste volume, higher production velocity, fewer defects at end of line testing (solder bridges, insufficient solder, tombstoning, etc.), fewer customer returns.
     If I look at the glass as half full I can see the good assignable causes. Variation is still the enemy, but if you know the underlying causes…go after them!

    0
    #167275

    Six Sigma Shooter
    Member

    Paulo,
    I think you are looking for something that doesn’t exist, at least to my knowledge.  A clean masking process was and probably still is a requirement.  I don’t think it can be made robust enough to ignore lint, oil or any other contaminant.  The timeframe of the example was circa 1993 / 1994, so things may have changed since then, but I rather doubt it.  I am not an expert in PCB manufacturing.  The point is that a special cause occurred, it was a good thing, and it led to changes in the process that greatly reduced the number of defects.  Sorry, I don’t have the answers you’re looking for. Maybe others that are more familiar with the PCB state of the art can give you more help in what you seek.
    Shooter

    0
    #167542

    New BB
    Participant

    You know I have been reading this site for several months now and I am amazed at the amount of bickering that goes on.I understand a disagreement but this stuff is nothing but mean spirited bantering.I thought this was a forum for enlightenment and cordial professionalism.I think some of us should re-read the etiquette guidelines.

    0
    #167545

    Fake ATI Alert
    Participant

    What  do  you  mean by : bickering, spirited bantering ?

    0
    #167749

    Dr. Scott
    Participant

    Real Data,
    Seems others have done the analysis of the data I posted for you, and have seen that there is such a thing as a good special cause.
    What is your delay? No analysis skills or pride?
    Dr. Scott

    0
    #167751

    Hulk Hogan
    Participant

    Hey dufus, if you want a fight, it’s not too early for a body slam.

    0
    #167756

    Dr. Scott
    Participant

    You have already been slammed. Now get up and contribute to this forum as you should, if the slam hasn’t disabled you.

    0
    #167780

    Hulk
    Participant

    Dr. Scott,Any good SPC text written including the first by Shewhart talked of
    patterns that were to your benefit and that the root cause of those
    should also be captured.Your announcement of having real data like you were in possession of
    something precious was just funny and showed how shallow your
    experience really is. Reminds me of all the SS consultants out there
    that have a single project they continually reference.

    0
    #167981

    6Sigma Below The Belt
    Participant

    Special causes depend on the situation. If you plot turnaround time in a control chart and have found 1 point below the lower limit, then that special cause is a good one, candidate for good practice BUT if it goes beyond an upper limit then that cause becomes bad.

    0
Viewing 58 posts - 1 through 58 (of 58 total)

The forum ‘General’ is closed to new topics and replies.