Quality News » Quality

Shewhart, Deming, and Six Sigma

QualityGuru — Tue, 15 Sep 2009 07:27:40 +0000

Another good article by Donald J. Wheeler
This article has focus on selected aspects of the work of Shewhart and Deming and how these compare with a common element of various six-sigma programs, a look at the concept of an operational definition, then turn to what it takes for improvement.
This will lead to a distinction between observational studies and experimental studies.
An Operational Definition

In the pre-publication drafts of Quality, Productivity, and Competitive Position Dr. Deming wrote:
“An operational definition consists of (1) a criterion to be applied to an object or a group of objects, (2) a test of compliance for the object or group, and (3) a decision rule for interpreting the test results as to whether the object or group is, or is not, in compliance.”

This definition closely parallels Dr. Shewhart’s opening statement for his (1939) book Statistical Method from the Viewpoint of Quality Control:
“Broadly speaking there are three steps in a quality control process: the specification of what is wanted, the production of things to satisfy the specification, and the inspection of the things produced to see if they satisfy the specification.”

This idea of an operational definition, which Shewhart and Deming popularized from the work of the philosopher C. I. Lewis, provided the seed for what grew into the Shewhart or PDSA Cycle.

Plan-Do-Study-Act Cycle does form a powerful framework for any improvement effort, it has often been reduced to a checklist to be followed mechanically. This has led to a proliferation of “expanded” PDSA cycles where each of the steps on the checklist are specified in ever increasing detail.
But before we go down this path, I would like to back up and generalize a bit.

In Dr. Deming’s own conversations, when individuals would start telling him about what they or their organization were planning to do, he would invariably have one of two responses for them:
“By what method?” or “How will you know?”
Either one of these questions would generally end the conversation since the individual would have no answer. After discerning this pattern to Dr. Deming’s responses, it finally occurred to me that these two questions corresponded to the last two parts of an operational definition. This realization, in turn, resulted in a generalization of an operational definition to become:

What do you want to accomplish?
By what method will you accomplish it?
How will you know when you have accomplished it?

Whatever you may be doing, until you can answer all three of these questions, you do not have an operational definition but merely a basis for an argument. Shewhart understood this, and in his work he used the concept of an operational definition in the development of the “operation of statistical control.”

Orginal and full text is on SPC Press online

Quality Management Practices in the Growing Telecom Industry – An Industrial Insight

QU-King — Fri, 12 Jun 2009 13:21:38 +0000

Quality Management has made significant inroads into Telecom industry worldwide. Initiatives like TL 9000 Telecom Quality Management System aim to provide “a consistent set of quality expectations to drive efficiency and performance” in telecom industry worldwide. Companies are adopting various quality tools in this sector and even the latest Quality Management initiatives like Six Sigma are being used.
Telecom industry is however in its growing stage and reliable data on adoption of Quality Management in this industry, particularly in developing economies, has not been drawn up. An extensive literature review has been done to explore the contributions addressing initiatives and issues in Quality Management in Telecom businesses. In parallel, Telecom business firms in Pakistan are investigated for their current and planned Quality Management Initiatives, including TL 9000.

The primary and expected outcomes of current and planned Quality Management programs and initiatives have also been explored. Telecom Industry in developing economies is attracting huge FDI (Foreign Direct Investment) in Pakistan and India, and in other developing countries; and the trend is expected to continue. The result of this study will contribute towards:

• Decision making knowledge required by current and potential investors.

• Provision of Industrial insight for practitioner and researchers worldwide.

• Addition to Sector Specific Quality Management Body of Knowledge.

Quality in telecom

The definition of “quality”, in the fledgling telecom industry in Pakistan, is at best obscure.

Quality is often defined imprecisely in textbooks (Allen 2006) in terms of a subjectively assessed performance level (P) of the unit in question and the expectations (E) that customers have for that unit. A rough formula for quality (Allen 2006) (Q) is:

Q=P/E
(1) Often, quality is considered in the context of many smaller parts of a larger service, such as the clarity of voice, connectivity, courtesy of operators etc, and the key issue is, why some fail to perform up to expectation and others succeed.

The small business and Total Quality Management

QualityGuru — Thu, 11 Jun 2009 20:08:27 +0000

One apparent advantage of being relatively small is that you needn’t bother with the big nostrums that gurus and consultants love to sell to big companies. Or is it an advantage? What if TQM (Total Quality Management), say, is the right road to making money?

In many businesses, top-class quality is actually becoming the only road. If you’re supplying giants (for example, in cars) they may well insist on quality certifications like BS5750. Getting these can be expensive and burdensome; they don’t require plunging into TQM – but if you’re going to all that trouble, it makes sense to launch a total quality drive.

Total quality made much sense to Jack McGavigan, just retired, in his 80s, as chairman of John McGavigan & Co, founded by his grandfather in 1860. The family firm was a traditional printer, but developed from the Sixties into specialists in ‘graphics related plastics technology.’ Touch switches and backlit fascia panels in cars are examples of the uses. McGavigan, rightly keen on innovation, has constantly expanded into new technologies and product lines. Consequently, the group’s car parts have a 12% world market share.

None of this might have happened if McGavigan hadn’t reacted to impending crisis in mid-1987. Costs were rising, the company was being squeezed between strong suppliers and stronger customers, and competition was intensifying. The firm had been using quality circles for years – with employees banding together to tackle quality issues, Japanese-style, in a voluntary, bottom-up effort to improve. But more was needed to avert disaster.

Quality was the answer, passionately advocated by Edward Smith, 47. Managing director of John McGavigan Automotive Products, he joined the company as a stripling in 1963. ‘Employees are the experts’, he says. ‘We wanted to exploit the potential of people,’ which demands ‘training, teams and communication.’
read full text Thinking Managers

Titan probe failure demonstrates pattern of quality control failures at NASA

QU-King — Sat, 06 Jun 2009 21:08:31 +0000

Here is a real whopper in the history of stupid space exploration tricks. The spacecraft that recently arrived on Titan was supposed to transmit images and experimental data on two data channels: channels A and B. But as it turns out, someone forgot to turn on channel A and so only channel B was available. And channel B was the least reliable channel of the two. It had slower transmission speed and was not designed to be the primary data channel.

As a result, some experiments were entirely destroyed. One scientist had been working for 18 years on an experiment to measure the winds on Titan, and his data was supposed to be transmitted on channel A. But since somebody at NASA or the Italian space agency forgot to turn on channel A, none of that data was transmitted. And thus, 18 years of waiting was lost.

All of this reminds me of the fiasco that happened when NASA sent the rovers to Mars. When the rovers first got to Mars and started taking pictures, NASA suddenly discovered that, ‘Gee! The onboard computer memory is getting full!’ For some reason, when you actually take photos, it takes up memory. And after only a couple of days, the rovers were unresponsive because they were constantly rebooting due to the fact that their memories were full. It was as if no one had thought to actually test the rovers here on Earth before launching them into space at a cost of 800 million dollars.

So I have a question. Why do we as a nation spend literally billions of dollars to launch hardware into space when that hardware is being programmed and operated by bumbling idiots who can’t flip a switch or remember to test the hardware before launch day?

Of course, NASA always manages to put a positive spin on these events. The Mars rover missions were ultimately proclaimed a great success, but only because some sharp-minded system administrators were able to ‘hack’ the rovers from Earth with a process that reprogrammed the way the rovers store images.

New Quality, Cost Analysis Tools Released

QualityGuru — Thu, 04 Jun 2009 07:01:21 +0000

Quality Control

Two small software vendors this week released business intelligence product upgrades that will allow manufacturers to more accurately and rapidly assess the cost and quality of their products.

SigmaQuest Inc., a provider of on-demand product quality analysis tools, released version 6.1 of its flagship SigmaSure product, which adds several new features, including support for identifying potentially costly, low-quality “maverick” products before they are shipped.

At the same time, aPriori, a provider of cost management analysis software, delivered version 5.5 of its self-named tool, which, among other things, allows manufacturers to compute the cost of a single component manufactured in multiple internal or supplier plants.

The most significant new feature in the 6.1 release of SigmaQuest’s SigmaSure is support for analysis of maverick parts and products, a capability that is increasingly important to manufacturers of electronics components, SigmaQuest’s largest customer base. A maverick part is defined by the JEDEC Solid State Technology Association as an “atypical product that exhibits anomalistic characteristics causing higher-than-normal levels of failure in the end-user application.”

Maverick parts, for example, may fall within prescribed functional test parameters but still fail often in the field, leading to costly product returns. Many electronic equipment OEMs, said SigmaQuest CEO Nader Fathi, have begun to require component suppliers to certify that they have analyzed part lots for their potential to exhibit maverick behavior.

“Just looking at product yields is not good enough to determine if products will behave correctly,” Fathi told Managing Automation today.

Resilient organizations

DrZoom — Mon, 01 Jun 2009 12:09:51 +0000

resilience

ISO9001, TQM, Lean, Six Sigma, Human Error Reduction, Quality, Environment, Safety, Health… are all ways of seeing (the same) things from a different perspective. Full of energy, we jump on every new hype or wave. Our ultimate goal: to create an ideal organization where nothing goes wrong.

But step by step, we start to realize that “zero fault tolerance” is out of reach, it’s an illusion. If we ever achieve to proactively track and tackle all the thinkable conditions that might influence the risk of error, some new conditions will loom up out of nothingness. We have to face the fact that we live in a very complex and dynamic world. Today’s reality will be completely out of date by tomorrow. Our organization … Read more on source