Six Sigma Thesis - Six Sigma Works: Fact or Fiction

[prototyper]

The major problem with 6S (IMHO) is that companies spend a fortune in training people to be green belts, black belts, pink belts with yellow spots , who can identify where the process variation is, but A) they then don't have the people who know how to fix the problem and B) it is far too late, as the real time to ensure minimal variation is at the product/process design stage.

I have seen several 6S projects and have yet to see significant improvements in terms of reduced variation or cost savings.

On the other hand, I have also seen some truly amazing results from Nissan, 2 day and 10 day, Kaizen activities. Working for a Tier 1 supplier in the early 90's, Nissan loaned us one of their trainers for a week as part of their supplier development programme. The first 2 projects resulted in cost savings of over £100,000 pa and measurable quality improvements.
The main difference is that the Nissan approach was to train multi-functional teams in the basic quality toolbox. Imagine how effective it is to have the people who really know the process :- operators, line technicians, etc., trained to perform effective brainstorming, ishikawa diagrams and to "think outside the box".

You can package these fads with buzzwords and mystique, but you can't beat giving the right people the right tools for the job!

[reynald]

In Reply to Parent Post by Paul Simpson

On the CQI web site there is an on line article on the merits (or otherwise) of six sigma. You can look at it here. Another restrained piece by me!

80:20!You gave it a really nice summary.

[Jim Wynne]

In Reply to Parent Post by Paul Simpson

On the CQI web site there is an on line article on the merits (or otherwise) of six sigma. You can look at it here. Another restrained piece by me!

Good job. Paul says,

If your 'belts' of whatever hue do not understand the basic concepts including common and special causes of variation then your six sigma programme is worthless.

That's it in a nutshell, I think. I've lately come to realize that the current fatuation with SS stems from neophytes recognizing that the coopted "core tools" of SS--understanding variation, finding root causes, methodological problem solving--are valid and valuable, and in their ignorance and zeal they take the plunge and accept the fallacies--the 1.5 shift, the folly of 3.4 DPMO, the ugly transmogrification of the simple PDCA into DMAIC--at face value. In other words, they see the tried-and-true assets of SS and assume that the garbage must be OK too. But strip away the garbage and the ridiculous caste system and you have relatively simple concepts that we've had for a long time.

[AlanC]

I use to work for toyota for many years and visited many suppliers who had or had not got 6 sigma controls. We generally were happy to work with those without and avoid those with in terms of developing them. The basics have already been said, that even those with had actually little knowledge of true root cause analysis and countermeasure. it has many branches which if applie correctly are very useful but you need many brances to make the tree, on there own the tools can fall down quickly. Similar to applying TPS and saying we do Kaizen etc etc. The use of the stats is very useful tools for proving c/m effectiveness, mintab is very good tool, but at Toyota such systems like having black belts, geen belts, SPC etc was once tried in a slightly different form but dropped early on since for the shopfloor and practical applcation it was too much to implement it as a working system, TPS tends to be on the surface much simpler, but actually also has its complications, the key point is everyone does it from top down

[qualityos]

Yellow Belts, Green Belts, and Black Belts .... Oh my!!

Just like Meaggie-Son said in the Karate Kid... a belt is only useful to hold up your pants. Karate lives in your heart and head. Same can be said with Six Sigma! The essence of Six Sigma is simply this... "To make the BEST decision you can based upon data and information". Organizations that get hung up upon how many "belts" they have trained have completely missed the point.

I am currently a Master Black Belt with over 8000 hour of instructing and leading experience in Six Sigma, Kepner Tregoe, and other problem solving methodologies. This is what I do for a living, and I love every minute of it.

There is a common theme throughout the entire problem solving / CI methodologies and that is to use FACTS, Not Opinions when making decisions.

I am still amazed to this day that most engineers and manufacturing folks don't use any kind of data to try to improve their process. They use perception, opinions, experience, gut feel, and what I like to call the "Poke and Hope" method (change something and cross your fingers and hope that it worked) to solve problems.

What is great about Six Sigma is that it provides you a road map for problem solving and enforces the discipline of making sound decisions based upon data.

Orgainazations that actually embrace the Six Sigma methodology and are disciplined enough to follow it, usually see significant results. Otherwise, it is just a rehash of TQM.

[Bob Bonville]

I was a legacy Honeywell employee for 30+ years. We routinely used many of the tools that the more formal 6 sigma program mandates, with success because we used them selectively based on unfavorable metrics that pointed to unprofitable, problematic areas of the business.

When Allied bought Honeywell and forced 6 sigma down everyones neck, it was overkill and marginallly successful.

Should 6S be utilized? yes!

Should it be utilized selectively? Absolutely, because you are expending valuable resources and you want to take advantage of these resources where and when they will provide you with the greatest benefit.

Bob

[artichoke]

In Reply to Parent Post by qualityos

Otherwise, it is just a rehash of TQM.

As Mikel Harry says, "Six Sigma is 80% TQM". It is unfortunate that the 20% Harry added is such nonsense !

Anyone who is serious about quality should forget the hoopla and go back to basics with Deming.

[bobdoering]

One of my favorite points when describing SPC for the uniform distribution is that is you want a practical, minimum number of operator interventions for adjusting a lean uniform distribution process, then set the control limits to 75% of the tolerance. If you want six sigma, then you need to set the tolerance at 50% of the tolerance. It will require more operator intervention, for which there is no economic benefit - and which the lean practitioner will wince.

If you are a lean six sigma practitioner, you will simply be torn.....