Reader Question on A3 Reports & Six Sigma DMAIC Framework

Here is a reader question for Professor Sobek:

Hello Dr. Sobek-

I have been tasked with developing and A3 Training and rollout program for my company. We have a need for a common communication and problem solving tool and to me A3’s are a perfect solution because of the thinking behind them. We need to foster a more rational and logical approach to problem solving.

Several of my colleagues think the A3 should be in DMAIC format vs. PDCA. What are your thoughts on how DMAIC fits within PDCA? To me DMAIC is just the Motorola coined version of the scientific method and PDCA is more fundamental. I have tried to build a case for PDCA as we are not doing Six Sigma here nor do we have any plans for rollout in the near future.

I would appreciate any thoughts on the matter.

Best Regards,

Professor Sobek’s response:

One question I have is where is the Act phase of the PDCA cycle in DMAIC? I suppose it can be fit in the Control phase or afterward but it feels like a forced fit to me…

Along the same vein one question I have with DMAIC is where do you fit the reflection and learning that we associate with the Act phase of PDCA. I asked this question to some master blackbelts at GE a year or so ago and their answers were enlightening. Most did not have an answer. The one that did said that his group added another step after the Control phase to essentially add an Act step…in other words something like a DMAIC-A.

The other thing that I caution people on is that PDCA cycles are frequently nested. You do a mini PDCA cycle for the “Do” step for example. I don’t know of DMAIC supports this kind of nuance? But the more important thing is what do you do with the learning from the Check step?

In talking with folks who are already trained about PDCA I find that many have a poor grasp of the Plan and Act phases and all the implied details. For example the Plan phase should ideally include (but is not limited to) the following:

  • First hand observation of the problem (Go to Genba)
  • Quantify the extent of the problem
  • If it is a process then draw a diagram of the process showing the problem(s)
  • Setting a goal or target
  • Root cause analysis
  • Generating effective countermeasures
  • Creating an implementation plan
  • Creating a follow up plan for confirmation
  • Discussing the above with related stakeholders

I can see where the DMAIC framework is fairly prescriptive of the key elements of the Plan phase and that has some utility. On the other hand the Act phase involves one of two directions depending upon the outcome of the Check step:
1) If you did not achieve the goal or target it means there was something about the problem or the current situation that you did not understand quite right. So you have to go back to the Plan step and repeat the cycle
2) If you DID achieve your goal or target it means that you have confirmed what you hypothesized is probably correct. In this case you want to share that learning by a) incorporating that learning into the daily work routines (e.g. standardized work, or Control from DMAIC) and b) communicating to others who might benefit from what you just learned.

“Act” is so important in PDCA because that is the step of the continuous improvement cycle that translates individual learning into organizational learning.  It’s what implies the learning! If you don’t do the Act step you’ve dampened the benefit from all that learning pretty significantly because it remains localized to one person or a small group.

Have you perhaps given this some thought: Come up with a practical problem solving method specifically for your company that addresses the best of PDCA and DMAIC for your unique context? It might be a way for you to address some of your colleagues concerns (to which I’m sympathetic actually), while keeping the critically important elements of PDCA that don’t seem to be emphasized in DMAIC?

Another thought: almost any reasonably thoughtful structured problem solving approach is going to be better than an unstructured approach. This has been pretty established in research. So getting something in place is far better than having nothing at all. Then apply continuous improvement to it as you gain experience using the method.

Good luck with your journey!

Regards,

Prof. Sobek

Comments are closed.