Continuous Improvement Techniques


Vero Flores

I work for a company, and we are implanted the qs9000 standard. The one section that is causing the most anxiety is continuous improvement techniques. I am unsure how to get started abd what to include. If anyone has ones insight that may help me, I appreciate it

[This message has been edited by Vero Flores (edited 09-06-98).]


Fully vaccinated are you?
There are many things companies do which address Continuous Improvement. Your nonconformance and corrective action systems are instrumental basics. A well done corrective action response could include Poke-Yoke / Mistake Proofing which are tools of continuous improvement.

Another input into continuous improvement would be acting on the FMEA. The FMEA is supposed to be a living document and the expectation is that a company have at least 2 to 3 items being addressed on the FMEA with some sort of improvement plans. Typically (and simplistically) the 3 top RPNs are to have some sort of improvement program.

If you do run charts (or use another statistical analysis technique) and track defects / defectives and act on problems, that is continuous improvement. Typically I see production meetings where trends are identified and improvement projects assigned.

The general guide is to take a look at your company and ask your self what you do to improve! What you do to identify and fix problems. Etc.

Scott Knutson

Look too for movement from manual systems to automated ones. Ex: if you track your product/service manually but are moving to a computerized system, or have just moved to said system, this too would constitute continuous improvement. Any effort that shows you have improved what it is you do, is continuous improvement

Jennifer Fry

Mark, I am a little confused with how you linked continuous improvement and tracking defects / defectives. I am questioning this just because of a note under 4.2.5 that states "...continous improvement is not possible until characteristics are conforming. If attribute data results do not equal zero defects, it is by definition nonconforming product...Improvements made in these situations are by definition corrective actions, not continuous improvement". Am I wrong then in assuming that continuous improvement, as associated with defective parts, can not happen until there are no defective parts?

Steven Sulkin

This is a good question, worth adding some breath to bring back to life. We are struggling with the same question. What differentiates corrective, preventive and continuous improvement. Our auditor explained it like this, corrective is in response to a problem found. Preventive is in anticipation of a future problem. Continuous is lowering the "level" of the bar. The continuous definition is a little screwy, but the intention is that actions are being taken not to respond to a problem, not to prevent it, but to improve the level of performance. I dont really get the importance of putting all this effort into distinguishing the three types. I just want to improve my competitive edge by getting better, faster than my competitor. I guess thats QS9000 for you. ;)

Kevin Mader

One of THE Original Covers!

You take corrective action to eliminate a problem that exists. By definition of corrective action, it requires the total elimination of the nonconformance and the future prevention of the nonconformance. Although I used the word prevention here it is not preventive action.

Preventive action is taken to correct a potential nonconformance that has not occurred. Since it has never occurred, you are preventing it from ever happening. A proactive approach. This is not to say that you can not use a methodology used in corrective action as a preventive action. For example, you find that 1 of 3 presses is producing nonconforming parts. You determine that the feeder was not feeding material fast enough to the machine. To take corrective action, you adjust the feeder so that the problem is corrected and can never occur again. Now the other 2 machines have been producing defect free. By adjusting the feeder in the same fashion, you have taken the corrective action for machine 1 and used it as preventive action for the other machines thus preventing any occurrences.

Continuous Improvement can either lower or raise the bar. It depends on which direction brings optimum results. Marc mentioned several good tools. I will use FMEA and Cost of Quality as my examples. FMEAs are performed to detect potential misuses, abuses of products or services. It detects the possibility of a nonconformance (Preventive Action). Here you are raising the bar for increased performance, reliability, quality, etc.. In calculating the Cost of Quality, you pick measurables in Prevention, Appraisal, and Failure costs. For Failure categories, and likely Appraisal, you would like to lower the bar and reduce costs. Now for something a bit screwy. You would like to raise the bar in Prevention, because as you know, an increase in Prevention reduces Failure costs. While these costs go higher, the trade off is the pay off in the savings for the Failure costs.

QS9000 is a decent set of requirements. A bit paper heavy on Customer records, but otherwise pointing organizations in a good direction. The real benefit is to learn and understand the CI tools and apply them companywide rather than just limiting them to the manufacturing environment.

System optimization occurs with the use of the three processes. Each has its place and I would have to say that I agree that categorizing each is not as important as getting the job done. It is important to know, however, how each is done, how it is recorded, and most importantly, how it is integrated into Management Review. I hope this helped a bit, but I think you have it pretty much right.

Don Winton

I am also intrigued by Jennifer's question as well. It is indeed good. However, since I am not QS savvy, but I will try to address from a Quality Management point of view.

Jennifer: "Am I wrong then in assuming that continuous improvement, as associated with defective parts, can not happen until there are no defective parts?"

Answer: Yes, that particular view is incorrect.

When I give a class in Quality Management, I have a standard question I always ask. "How do you get a car to travel at the same speed?" Typical answers vary, usually along the lines of "cruise control." The solution: Park It. The follow-up question is "How do you produce a perfect product (read zero defects)?" The answer: You Produce Nothing (theory of common cause variation). Then I go on to explain that continuous improvement does not mean "perfect" or "zero defects." It means taking a process that is producing at some level of quality and making it better (higher or lower, depending on the case in hand, as Kevin stated). Not perfect, better. Then you do it again, and again, and again. You approach perfect, but you do not obtain it. Continuous is just that, continuous.

The inverse to continuous improvement I also try to teach is the concept of entropy (From Slater's book). I give it a face (usually very mean and very ugly) and explain it this way:

"Entropy has one function in life. His job is to make things just a little worse today than they were yesterday. The clutter on your desk, the typical kitchen 'junk' drawer, etc. Given time, a little worse adds up to a LOT worse." I give the example of a car. The class is told, "Right now, even as we speak, your car is rusting. You may not see it today, tomorrow or even next week. But, it IS happening. Ten or twenty years from now, it will show up. What do you do to stop it? You wash and wax and wash and wax and wash and wax. Now, apply the same principle to your job. What can YOU do to make things a little better today than they were yesterday? Entropy is a tireless and fearless warrior. He will never stop. You cannot win the war, but you can win battles. YOU must be better. What can you do to stop HIM!"

Anyway, these are my views on continuous improvement. Hope someone can use them, or did I just muddy the water.


[This message has been edited by Don Winton (edited 11-20-98).]

Kevin Mader

One of THE Original Covers!

You bring up a good point (as you normaly do, no mud in my eyes) about "Zero Defects", perfection, and continuous improvement. People become so preoccupied with definitions of exactness that the bulk of the concept is missed. Zero Defects, for example, becomes such an intangible and unattainable goal. It is impossible, so why try to get the impossible? I feel that this is the case because management does not sell the concept correctly. Zero Defects is sold as a destination rather than a direction. It is the direction and the unattainable goal of Zero Defects that creates the continuous improvement journey. A journey towards perfection, but not perfection. This is a good topic for the TQM forum.

QS9000 has the advantage on paper over ISO9000, atleast for now, in that it requires Continuous Improvement. Both, however, require 'effectively' accomplishing the standard requirements which can be construed as continuous improvement. That is what I teach in my organization.

Back to the group.....


Fully vaccinated are you?
Not a thing I can add. All of you have answered in great detail and quite eloquently.

Thanks to everyone. Your participation is what makes these forums valuable to all of us!

Leslie Garon

Only one thing to add,

Percision = Exactness
Accuracy = Proximity, within parameters

Remember what your goal is.
Top Bottom