The Poka Yoke

Editor's Note: This song was sent to us in response to the articles on lean in the April 2003 issue.

Sung to the tune of "Hokey Pokey"

You take the waste away,
And keep the kanbans full.
You standardize today,
And use a supermarket pull.

You do the poka yoke,
And you turn your plant around.
That's what it's all about.

Change your dies real quick,
And watch your time for takt.
The current state is sick;
Senior management must act.

You do the poka yoke,
And you turn your plant around.
That's what it's all about.

Turn on your andon light
Increase your OEE.*
Just-in-time is right,
And inventory's never free.

You do the poka yoke,
And you turn your plant around.
That's what it's all about.

So do a kaizen blitz,
And keep your work cells clean.
Ensure the flow rate fits,
And map your value stream.

You do the poka yoke,
And you turn your plant around.
That's what it's all about.

Livonia, MI

* OEE = operating equipment effectivity.


Case Study Provides Much Needed Jump-Start

The article "Eastman Chemical's Success Story" by Gary Hallen and Robert J. Latino was quite timely (June 2003, p. 50). I am a quality assurance manager who heads a project team in charge of addressing the reduction of customer complaints.

We have implemented some changes with some improvement, but I think root cause analysis training and the use of logic tree basics will provide the jump-start we need to achieve our goal. I've shared the article with my project team and plan to recruit some assistance in this area.

Thank you for a great story!

WIKA Instrument
Lawrenceville, GA


Any Customer Input Is an Opportunity

While I agree with the idea of using logic tree analysis to determine the causes of customer complaints, I don't agree with using it to reduce the number of customer complaints. This can foster a culture of not cataloging customer complaints. The real goal should be to get to the root cause of the complaint.

A culture should try to obtain as much customer input as possible. Any customer input, whether it is actually a complaint, is an opportunity.

Grand Island, NY


Does Eastman Measure Customer Satisfaction?

"Eastman Chemical's Success Story" describes the methods the company uses to deal with customer complaints, but I wonder if the company has a business process to measure customer satisfaction because it is not mentioned in the article. Most quality frameworks, including ISO 9000 and Baldrige, indicate a "measure customer satisfaction" business process is necessary.

Obviously, complaint measurement and resolution should be done. However, a customer satisfaction business process is also required because it involves more than just the measurement, reduction and resolution of complaints.

If Eastman Chemical does not measure customer satisfaction, how does it answer the questions on customer satisfaction when audited, assuming it is ISO 9000:2000 certified?

If the company does have such a process, what kind of survey results were measured regarding customer satisfaction and dissatisfaction? If collected, did the satisfaction measurement correlate with the indicated level of customer complaints? And, moving toward best practice, what kinds of permanent anticipatory or preventive actions were or are being taken regarding customer satisfaction?

Eden Prairie, MN

Author's Response: Yes, Eastman Chemical has had a comprehensive process for measuring and managing customer satisfaction for many years. As a 1993 Baldrige Award winner, the company has long found this process vital in the management of its total business. This process has also been instrumental in the ISO 9000 registration of all its major facilities.

The results achieved in using root cause analysis to significantly eliminate causes of customer complaints were largely enabled by the quality and customer oriented culture already within the company.

Eastman Chemical's customer satisfaction process has long involved a multitude of methods: a structured customer survey process, a complaint handling process, a lost-and-gained business process, verification by outside consultants, participation in formal benchmarking studies, audits by customers and quality partnerships.

Customer complaints measurement is only one piece of the more comprehensive customer satisfaction process employed by Eastman Chemical.

Eastman Chemical
Kingsport, TN


The Data Duplication Demon Strikes Again!

In November 1997, I was placed in the position of document control coordinator at my facility. Mikel E. Janitz's "One Good Idea" column ("Document Control," June 2003, p. 104) didn't just come close to describing our current system, it nailed it on the head. Bravo!

But what did Janitz mean when he said, "Provide for multiple access points (input/output)"?

Also, control of electronic documents cannot be complete without control of internal e-mailing systems. As soon as a file is e-mailed to one recipient, three electronic copies exist: the original, the copy attached to the sender's e-mail and the copy attached to the recipient's e-mail. There is now a two out of three chance one of the nonoriginal copies will be revised.

This is further complicated by the almighty signature-designator of the mastered hard copy. When electronic signature capabilities are far into the future, control of distribution will be unavoidable.

In the automotive industry, production critical data must be reviewed and approved by multiple parties. So the recipient, who happens to be one of the approvers, revises his copy of the e-mail, prints it out, has the others review, approve and sign it. Voilà, the demon of data source duplication has once again possessed another poor soul!

Associated Spring
Corry, PA

Author's Response: I'm glad you agree with our (document) information control approach. To answer your question, multiple access comprises two things: a company webpage and a common drive where the working documents are stored. The common drive is split into three folders:

  • Drafts.
  • Procedures.
  • Intranet.

All folders are secured by read only access through IS, except for those employees who have authority to create, update and approve. Documents in the procedures folder are password protected. Access to the intranet folder is limited to three people, and those files are also password protected.

Anyone can pull up a document in the draft folder, read it and do a save as. We also e-mail files, but it is known that only information on the intranet site is current and accurate. Hard copies are considered obsolete. As for approvals, the typed name is the electronic signature. Nothing fancy there. There has to be some trust in the system.

There are computers throughout the plant, in production, offices and remote locations, all of which provide access to the common drive and intranet site. But only three people can actually make changes that are reflected on the intranet for all to read.

If I can comment on your e-mail issue, we had the same problem. We don't want hard copies signed and then have to work to control them. By our definition, hard copies are suspect and considered uncontrolled. So remove the requirement of electronic signature or hard copy signatures. Keep it simple, and keep saying it is the information you want to control, not the document or its form.

Mercury Marine-MerCruiser
Stillwater, OK


More Variety Needed In 'The Face of Quality'

I really like the changes the QP team is bringing to the magazine. It is easier to read, and the updated format works well.

I would like to suggest, however, that you consider featuring a larger variety of quality professionals in "The Face of Quality" section in "Keeping Current." How about featuring a longtime quality inspector or auditor? Or a quality engineer who has been making a difference?

Seeing profiles of award achievers, directors of this and professors of that is good, and I do find those interesting to read. I would just like to see a balance.

Again, thank you for the changes. Overall they are good and refreshing.

Northrop Grumman

Editor's Response: Thank you for your feedback. We profiled an inspector in the past, and we always welcome suggestions of names of people to profile. You can send them to sdaniels@asq.org.


A Dry Subject Made Interesting To Read

The supply chain management article (Peter E. Pylipow, "My Supplier's Capability Is What?" May 2003, p. 60) was great! This is coming from someone who works in marketing at a small community college and who, up until now, didn't really know what supply chain management was. Thanks for proving seemingly dry subjects can still be presented in an interesting fashion. I will be saving this piece as a prime example.



Method May Be Statistically Invalid

I would like to congratulate Peter Pylipow on his narrative style and the simple manner in which he presented statistical topics.

I agree bundling data from different cavities and presenting them as one population will not give you a normally distributed representation of the process, and therefore, any Cpk or parts per million values obtained are not relevant. Nevertheless, most customers demand an overall process capability and only one Cpk value for the whole process.

I would like feedback regarding the following method of obtaining an overall Cpk value for the whole process, using individual cavity capability data:

1. Carry out the process capability for each cavity.

2. Obtain the Cpk value, z-value and defects per parts per million (DPPM) rate.

3. Average out the DPPM for all cavities (as recommended by Pylipow), thus obtaining the overall process DPPM.

4. Find the equivalent z-value for this DPPM.

5. Calculate the overall process Cpk (Cpk = z/3).

I've used this method, and it looks as if it's a good representation of the overall process capability. At least it's more realistic than probability data obtained from one capability study of bundled cavity data.

However, I'm concerned about step four. When the DPPM rate is transformed to z, the assumption of an overall normal process is being made once more. Does this invalidate the statistical relevance of this method?

FNGP Cleveland, GA Facility
Gainsville, GA

Author's Response: Your method is exactly right, and we've used it here at Eastman Kodak in the past. We called it "back-calculating the Cpk" or "equivalent Cpk." You can call it whatever you'd like--perhaps the "Bonello Cpk." It's probably in a book somewhere, as well.

Don't worry about step four. You are just calculating an equivalent, and as long as the people you discuss it with understand its origin, you are not at risk of invalidating statistical relevance.



Contact 'QP'
We welcome your letters. Send them to EDITOR, ASQ/QUALITY PROGRESS, 600 N. PLANKINTON AVE., MILWAUKEE, WI 53203-3005; or e-mail them to editor@asq.org. Please include address, daytime phone number and e-mail address. All letters will be published on the QP Discussion Board, or you can post your comment on the QP Discussion Board directly at www.asq.org. We reserve the right to edit letters for space and clarity. Opinions expressed are those of the writers and not necessarily of ASQ or Quality Progress.

Average Rating


Out of 0 Ratings
Rate this article

Add Comments

View comments
Comments FAQ

Featured advertisers