Home > The Lean Post> CI Sustainment: A Hiding Place for Complexity
The Lean Post
Sharing how the world is making things better through lean.

CI Sustainment: A Hiding Place for Complexity

by Lynn Kelley
March 30, 2016

CI Sustainment: A Hiding Place for Complexity

by Lynn Kelley
March 30, 2016 | Comments (8)

Have you ever drafted a continuous improvement plan and implemented it, only to feel later on that you may have made it more complex than it had to be? It’s not uncommon. And chances are you might have been right.

Whenever I think about this, I think back to many years ago when W. Edwards Deming started to teach us that variability was bad. Nowadays, it’s an automatic response: “Yes, variability is bad.” Before Deming’s lesson, there was a time when I would introduce continuous improvement concepts to groups of people – and most would say that variability was good. I would ask for examples of “good variability,” which would quickly bring the group to the epiphany that variability is actually bad. We’ve come a long way in terms of understanding variability. I will share more thoughts on variety (generally good) and variability (generally bad) in my next post.

But on to complexity. Today, I think we’re at a similar place with our understanding of complexity. How many times has your continuous improvement solution wound up being more complex than the original problem? The Seven Wastes and poka yoke tackle complexity from a building-block aspect, in that when applied to the problem at hand, we should theoretically be removing steps, reducing time, and simplifying processes. I think the real problem comes when we fail to fully apply these tools directly to our lean solutions, rather than just our initial examination of the process.

This is especially prevalent in improvement activities that involve defect prevention and quality assurance. If a clear poka-yoke isn’t evident, we tend to build in additional inspection steps, complexity and waste that may solve the defect problem, but nonetheless adds even more complexity to the process. Additionally, when teams place a focus on sustaining improvements, there is sometimes a tremendous amount of complexity added to the back-end of the process to ensure that it sustains. Generally this is in the form of inspections and audits. When we add up all of this good-intentioned complexity, our solutions are in danger of overburdening the organization, creating a lack of flexibility and nimbleness.   

It’s a delicate balance, but sometimes we err on the side of adding complexity when faced with these situations. I suggest an additional pushback when you are ready to implement a lean solution. One final check to see if there’s a way you can simplify it even more. It’s true that by the time the CI plan is ready to launch, the teams who made it are often ready to rejoice in their imminent successes and don’t have the energy for one more look. But sometimes, one more look can make all the difference in the long run.

Have you seen a rise in complexity/bureaucracy in your organization, even as you're making progress in your continuous improvement activities? Do you think it's partially due to the back-end sustainment complexity of your CI activities?

The views expressed in this post do not necessarily represent the views or policies of The Lean Enterprise Institute.
Search Posts:
February 22, 2018 | 3 Comments
October 24, 2017 | 8 Comments
Was this post... Click all that apply
HELPFUL
9 people say YES
INTERESTING
16 people say YES
INSPIRING
4 people say YES
ACCURATE
10 people say YES
8 Comments | Post a Comment
Mark Graban March 30, 2016
2 People AGREE with this comment

Dr. Deming talked about "variation." Is that the same thing as "variability?"

I don't think all process variation is necessarily bad. One surgeon is left handed and one surgeon is naturally right handed. They each do a surgical procedure with their dominant hand. That's "variation" in the process that probably doesn't matter for outcomes. As a patient, I wouldn't want my left-handed surgeon to be forced to use his or her right hand.

Dr. Deming really also talked about understanding variation in data and metrics and the need to understand common cause variation vs special cause variation, which is a different issue than "are pepole doing the work the same way."

Reply »

Lynn Kelley March 30, 2016
2 People AGREE with this reply

It's great that you have identified that Deming called it "variation."    One of my first introductions to CI was as a member of a "Deming Study Group" in Detroit, back in the 80s.  To be honest, every time I think about variation being bad, I think I'm quoting Deming, but I didn't fact check to verify that the actual word was correct.   Thanks for identifying this disconnect.    I identify variation and variability as beng the same thing, but now you've given me something to think about.

 

I have also spent a lot of time thinking about "good variation" and "bad variation" because of the point you have mentioned.   My next post on this subject addresses exactly that.   Spoiler Alert:   I believe what we call "good variation/variability" is actually variety.   I look forward to your feedback when this topic posts.

Reply »

Herbert Gensch March 30, 2016
2 People AGREE with this reply

Variation is bad...perhaps unnecessary variation is bad.  "Everything is a process and all processes exhibit variation."  It is the variation beyond what customers, internal and external, expect we should, and must, correct.  Better yet, design the product/process to contain variation much less than what all desire.

Reply »

Mark Graban March 30, 2016

Herbert - I agree with you. There's some process variation that honestly doesn't make a difference and some variation that has a HUGE difference.

Mark Graban March 30, 2016
1 Person AGREES with this reply

Lynn - I'll look forward to that.

BTW, my first exposure to Deming (before anything Lean/TPS) was when my dad, a 40-year GM engineer, attended a Deming four-day seminar. 

I'm curious to think about the difference between process variation and results variation.

SPC charts help us see variation in results... what's caused by common causes (which could be daily variation in process and methods) and special cause.

Deming also warned that "tampering" with a system can increase variation in results, as seen in the "funnel experiment." Sometimes fussing around with the process is improvement, sometimes it's just tampering. Hard to know when sometimes until we try.

Reply »

Robert Miller March 30, 2016

A great tool to drive variability of of a process is Stable Operations, It measures performance helps you identify resons for not making targets (dent reporting) and identifies opportunities for beating targets. It can be used to identify variables like waste and performance / trainnig opportunities.

Reply »

Lynn Kelley March 30, 2016

Quick reply to Herbert - Yes!  Thanks for pointing out that driving out all variability (beyond what the customer wants/needs/expects) can also add unnecessary cost.  I'm not sure if it would fit into the "overprocessing" category of waste--but it is some kind of waste.   This could also dovetail with Mark's comment of overtampering.

 

Quick question for Robert - I don't know of a specific tool called Stable Operations (beyond knowing that's one of our key goals in operations).   Is that something everyone knows about and I'm just in the dark or is that a tool that isn't used often?   What does the tool look like?

Reply »

Bill Waddell March 30, 2016
1 Person AGREES with this comment

There is no such thing as 'good variability' or 'good variation'.  Variation is the measure of how well (or not) the process is under control.  It is fundamental that the lowest variation process results in the shortest cycle time process; and the shortest cycle time process is the best cost and quality process.

Best example:  Henry Ford.  Any color so long as it is black meant no variation, resulting in the shortest cycle time auto manufacturing, resulting in extrememly low cost and high quality.

As far as the left-handed and right-handed surgeons are concerned I believe you are looking at two different processes:  left-handed surgery and right-handed surgery; and each of those processes are best performed with the least variation.

Reply »

Search Posts:
February 22, 2018 | 3 Comments
October 24, 2017 | 8 Comments