Standard Lean Logic Flaw

This is a cross posting from over at the Lead Edge website. The Lean Global Networks asks a question about changing standards and having to re-train everyone. This implies improvement is limited by the ability to conduct training.

This question unfortunately reminds me of the old adage in problem solving that vague fuzzy problem statements lead to poor causal analysis and then in turn poor countermeasure selection space. Any results will usually be limited in nature if they are evident at all. In this post I will point out some problems induced by the above definition of standards, the flaws in the logic at least with respect to actual TPS practices, and what I instead suggest. I will try and utilize some generic examples for contemplation and clarification.

The question above in this month’s LE question implies that the primary definition of standards is “the best way known to perform a certain task”. Indeed this is the normal Lean definition or answer I see supplied but it is only a partial answer at best and a potential mistake at worst.

Let’s look at an alternative dictionary definition of “standards” and one that is closer to the TPS meaning of standards in principle.

The important operating definition for standards in this case is to function “as a basis for comparison” or as a “principle that is used as a basis for judgment”.  See if you can first grasp the key difference on your own before reading on…

Fujio Cho the former President of Toyota often made this point in internal lectures and operational reviews all the time. “Show me the standard by which you are comparing the previous state to the current one”.  In other words if you did not have a standard then there is thus no quantitative basis for comparison to occur and no true improvement can be measured or claimed. This reality is quite a problem that plagues the lean movement which tends to require posting of textual visual standards (in the narrow form of standardized work) and other documents in the naïve belief this will solve most all problems. That sort of thinking is not working out so well in reality as companies are finding out. I often refer to this as the Lean Wall Paper Problem.

However other than the problematic definition implied in the question this month I also intensely dislike the underlying logic exhibited in terms of the follow on statements. For example if standards are the best way to do things, then training and re-training is the automatic preferred solution, etc. That is a nice conclusion for internal and external consultants as it would essentially create a never ending stream of work for those parties! Perhaps this is the way they have been taught to think. Unfortunately this is lacking in terms of critical thinking and far removed from actual best TPS practices.

As I implied in the opening paragraph poorly thought out problem statements or questions lead to poor analysis and in turn poor countermeasures. In this case the implied remedy or improvement is “re-training” which automatically means a very limited and human centric solution space is being considered.  In other words this is the result of what my friend Tom Harada decades ago called the “5 Who’s” of analysis rather than the “5 Why’s”. He is also partial to the “5 How’s” but that is another story. Alternatively if you prefer this logic above regarding standards and training is analogous to the red bead exercise conducted by W. Edwards Deming years ago which highlighted how the operator is not the sole determining factor in the process and often improvements can’t be made alone by manual methods or exhortations to perform better. Or by required “training and re-training” for that matter either…

For clarification let’s use an example that is not the simplistic manual operator case which is overly used or extrapolated to all cases in Lean. A drilling machine (everyone can envision a drill?) makes a hole in a part for some purpose. The quality standard calls for a hole diameter dimension of 50.0 millimeters +/- 0.050 millimeters. Let’s just ignore depth for simplicity. The machine initially worked fine for quite a few years. It was measured to reliably perform to a standard of 1.33 Cpk in terms of process capability. In terms of maintainability the slide unit housing the drilling head was measured to a vertical travel standard of 0.010 millimeters over a 10 centimeter distance. Furthermore the spindle head rotating the drill unit was measured to perform to a 0.005 millimeter run out standard. Then after a while one day the machine stopped being able to perform to these standards and also no longer could achieve quality requirements. I could go on with more but this should be sufficient.

After extended problem solving and consultation with the machine builder the bearings were  ascertained to be worn out and changed to a newly made and different type that was deemed to be better for the type of work (drilling into a specialty material) being performed.  So instead of merely replacing the old ones new bearings were installed and suddenly the machine again worked as before. Indeed the machine not only met but it exceeded all previously measured standards from before. Of course proper measurement follow up over time will be the judge if it lasts longer and continues to perform better than the previous type. From the operator point of view no “training” is required. Workers operate the machine as before, maintenance performs work as before, quality measures product as before. Daily management goes on however quality and now output, etc. is better. Engineering has follow up work and corporate standards will be updated to the new type of bearings in this application for future purposes pending verification, etc. This is a good example of highlighting what is wrong with the above question and training implications listed in the original question.

Here is another example of lean standards versus actual TPS thinking. Production volumes and types proliferate over time in a facility driving more and more change overs. A machine which once functioned fine is now a bottleneck and source of major downtime in the plant. It is driving overtime and poor morale as the changeover work is difficult to perform and yet is increasing all the time. The Lean solution is to “standardize” the process based upon whoever performs it best and rely upon that method for training or re-training everyone else. This answer unfortunately  is also analogous to the “5 Who’s” as well and simply results in training and re-training in an attempt to get more people to perform to the level of the best person. That is an uphill battle whether you use standards, standardized work, or job instruction training. Also even that is not enough in this case to solve the current magnitude of the problem.

A big problem in this scenario is that no improvement attention was applied to the actual machine nor was proper TPS thinking applied. The solution space converged simply to the machine operators and their method. A better solution would use automatic alignment aids and work simplification that makes the work radically easier, safer, and faster to perform.  The task would be simplified and improved instead of merely standardized. An even better solution however would eliminate the need for the difficult changeover work in the first place by more fundamental process improvement. Product standardization and process engineering would be part of the solution space of course. However please notice that the end result in this case is the not the presumed endless drafting of better “standards” or “re-training of personnel”.  That is lazy thinking and missing the fundamental improvement spirit of TPS. The proper TPS solution space in this case involves investigation of options that eliminate those auxiliary wastes as well. Be suspicious of any thinking process that primarily ends up with re-drafting standardized work, re-training employees, as the solution space. Normally it is missing the more fundamental cause or problematic area…That is a continual problem I see with Lean programs.

You might think I am being overly rigid here or narrow minded. However I no longer view “Lean” as the same thing as “TPS” from my experience. Perhaps this is fine. Perhaps Lean will become something better. However based upon my observations and interactions with companies practicing Lean I highly doubt it. The data from the Shingo Prize Institute regarding companies attempting Lean appears to bear out my pessimism as well unfortunately.  The quality of the thinking patterns I observe in “Lean” programs is often suspect compared to my actual experiences in Toyota. Answers that involve mandatory map drawing, standards posting, and training, etc. are often examples of very superficial thinking and often failure to consider what the real problem is…

In this case I think a better set of questions which are more reflective of actual deeper TPS thinking are the following:

  • What exactly is the process in question first of all?
  • What are we defining as the relevant standard in this case? (Hint it usually should not be standardized work).
  • What is the actual current standard? (quantitative not map or text or pictures)
  • How exactly is it measured?
  • What is actual performance compared to standard? (A gap indicates a problem).
  • What is the root cause of the gap to standard?
  • How is the work performed? (not just an operator time study!)
  • What can be eliminated, combined, rearranged, or simplified? (not just standardized)
  • Or if there is no gap then how can you improve the standard? (Think entire solution space of time, cost, quality, safety, productivity, man, material, machine, methods, product design, process design, etc. not just training).

These questions and thoughts when properly applied will lead to a better solution space or root cause insight than simply drafting better operator work standards and re-training people. Employees are getting fed up with Lean Wall Paper. I can’t say that I blame them either. The problem lies in faulty thinking patterns and the inability to come up with better solutions.