hi,
in my estimation a process has to be 'under control' before you can perform designed experiments to improve the process. otherwise you are wasting a lot of time and money. a process may be lousy, but it needs to be consistently lousy before you make doe attempts to improve it.
phil
-----Original Message-----
From: Mark Orlowski [mailto:[log in to unmask]]
Sent: Tuesday, January 08, 2002 7:26 AM
To: [log in to unmask]
Subject: Re: [TN] A SERIOUS QUALITY QUESTION
Lou,
Good quotes. I think the Deming quote pretty much covers it all.
But on the "everyone knows", meaning "not everyone knows" front:
Almost everyone in the quality field repeats the mantra: the process must
be in control (stable) before you can describe its capability (calculate
Cpk, Cp, etc.). I'll continue to voice my dissent, it just ain't so.
One doesn't need very sophisticated methods to identify special cause
variation from common cause variation in a data set. Just plotting the
data and using a wizened old critical eye is often enough. Then there are
simple methods to filter this special, assignable cause variation out of
the calculations for Cp and Cpk. So you CAN derive a valid Cpk from an
unstable process.
Of course, a Cpk has no predictive value unless your process is brought
under control. And lots of Cpk's are calculated by folks who don't
understand variability and/or their process. And Cpk's are subject to
manipulation (fraud), like you said. So maybe it's best to stick to the
dogma that the process must first be in control. But it ain't so.
Final editorial: we place too much emphasis on indices like Cpk when we
consider process capability. It's just a number.
On Mon, 7 Jan 2002 12:14:30 -0500, Lou Hart <[log in to unmask]> wrote:
>Earl, let me offer 3 quotes.
>
>The first I don't remember exactly, but everyone's heard it. (I say
>"everyone", having told others any number of times "When I hear, 'Everyone
>knows', I hear 'Not everyone knows'"). The one where Lord Kelvin says when
>you can express your knowledge in numbers, you reached the stage of science
>- or something like that.
>
>"Understand variability", Deming.
>
>"The key to success in engineering is understanding the variability in the
>measurement process, and using that understanding to control the
>variability in the manufacturing process", Hart. It could be said better.
>
>Referring to your posting below, I would say that a process is in control
>if it is consistent (stable) - its mean and variance are (to a chosen
>statistical confidence) constant. In-control processes can produce lots of
>bad product. Before determining a process is or is not in control, some
>knowledge of the variability in the measuring process is needed.
>
>Cpk is what I would call process performance index - a number. Cp is
>process capability. Process must be in control before calculating these
>indices. These indices depend on spec values, so are subject to
>manipulation, especially if engineering, sales, management doesn't really
>know what it wants.
>
>Let me try to adopt the viewpoint of designers in the DFM picture. Is
>manufacturing capable of providing technical guidance to designers with
>regard to manufacturability, based on data and engineering rather than
>speculation? In many cases, I suspect not. Maybe that is why it DFM seems
>to take place so infrequently. If data and engineering are lacking, the
>two groups may collaborate to generate them. Would that collaboration be
>part of DFM? Lou Hart
>
>-----Original Message-----
>From: Earl Moon [SMTP:[log in to unmask]]
>Sent: Monday, January 07, 2002 11:00 AM
>To: [log in to unmask]
>Subject: [TN] A SERIOUS QUALITY QUESTION
>
>Kind folks,
>
>Before I go off in the frozen tundra in my current MI location, to finish
>my
>book, I need some input on the following:
>
>Process control is defined as using process capabilities
>in an efficient, effective, consistent manner. To say a process is
>in control means the process is operating without assignable cause.
>If assignable cause is found, the process is out of control, and
>must be corrected, redesigned, and re-characterized. If a process is
>found to be in control it does not mean it is capable of meeting
>product acceptance specifications. It simply means it is consistent ? good
>or bad.
>
>Process capabilities charts (Cpk) describe the relationship between process
>control and product quality. When process capabilities are statistically
>proven to provide specified product quality, they are capable of providing
>product meeting customer/contract requirements.
>
>The above factors should be considered at the design concept level. Then,
>DFM/CE has a firm footing so a concept may be turned into a design capable
>of being manufactured as a product meeting specified customer contract
>requirements.
>
>Just a bit wordy, as usual, but do these words have validity in today?s
>?just do it? world. I mean, does what Juran, and others, have been
>preaching
>so long, have meaning for us in light of some of the previous discussions
>concerning quality systems, ISO, too much restraint, not enough creative
>freedom, and all the rest?
>
>MoonMan
>
>------------------------------------------------------------------------
>---------
>Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
>To unsubscribe, send a message to [log in to unmask] with following text in
>the BODY (NOT the subject field): SIGNOFF Technet
>To temporarily halt delivery of Technet send e-mail to [log in to unmask]:
>SET Technet NOMAIL
>To recieve ONE mailing per day of all the posts: send e-mail to
>[log in to unmask]: SET Technet Digest
>Search previous postings at: www.ipc.org > On-Line Resources & Databases >
>E-mail Archives
>Please visit IPC web site (http://www.ipc.org/html/forum.htm) for
>additional
>information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700
>ext.5315
>------------------------------------------------------------------------
>---------
>
>---------------------------------------------------------------------------
------
>Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
>To unsubscribe, send a message to [log in to unmask] with following text in
>the BODY (NOT the subject field): SIGNOFF Technet
>To temporarily halt delivery of Technet send e-mail to [log in to unmask]:
SET Technet NOMAIL
>To recieve ONE mailing per day of all the posts: send e-mail to
[log in to unmask]: SET Technet Digest
>Search previous postings at: www.ipc.org > On-Line Resources & Databases >
E-mail Archives
>Please visit IPC web site (http://www.ipc.org/html/forum.htm) for
additional
>information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700
ext.5315
>---------------------------------------------------------------------------
------
---------------------------------------------------------------------------------
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL
To recieve ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------
---------------------------------------------------------------------------------
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL
To recieve ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------
|
