I read with interest the April issue of QP and the many articles regarding the quality problems Toyota is having. I saw very little discussion regarding quality of design, which appears to be the root cause of most of Toyota’s current quality problems, including the stability-control issues of its Lexus GX 460 premium SUV.
The Toyota Production System (TPS) focuses on conformance to the design, not determining that the design is fit for use and foolproof. What the TPS guarantees is that when a design deficiency exists, all of the products will have the design deficiency.
As a solution, several writers mentioned a robust system for customer feedback. While I support such systems, they are reactive in nature. If the manufacturer uses the system to identify design deficiencies, the lag time is extreme. It is best suited for addressing the softer areas of customer dissatisfaction, not to identify significant product deficiencies.
Having worked in the automotive and aerospace industries, I can assure you the former has much to learn from the latter when it comes to ensuring design quality. This raises the topic of software design, redundancy and quality assurance.
As automobiles become computers with wheels, it behooves the automotive industry to implement a robust software quality assurance program. The industry can benefit by studying and adopting the standards and work of the International Aerospace Quality Group, which addresses design quality with the same intensity as conformance.
I am writing to comment on Henry Lindborg’s Career Corner column, "Lesson Learned," in the April 2010 issue. Toyota’s troubles are significant for quality professionals, but not in the way you might think. Quality at Toyota, in the form of the infamous TPS, did not cause the current problems at the automaker.
What appears to have happened is that Toyota changed its strategy from making the best (highest-quality) cars in the world to being the biggest automaker in the world. In the process, it unintentionally drove home a different set of cultural behaviors that resulted in the woes that played out so publicly in recent months.
How did this happen? As business consultant and author Jim Collins points out in many of his books, it is one thing to achieve a certain level of organizational excellence and another to maintain it over time. It is a characteristically human trait to relax after achieving a goal.
It may be that after years of focusing on quality as a strategy to differentiate its product and drive growth, Toyota executives felt they had mastered quality and could move on to the strategy of being the largest automaker. In doing so, they reduced quality to an operational—or worse, a tactical—initiative within the organization. The unintended consequence of such a change in strategy is that people may do things under a strategy of "be the biggest" that they may not do under a strategy of "be the best."
Toyota made its reputation on quality, and that is what most people remember. Sadly, the implication of this for quality professionals may be the erroneous connection in people’s minds that quality caused the current problems—or did not uncover them—or that Toyota no longer makes a quality product.
Clearly, something changed in the system. And while quality—in the form of the TPS and the Toyota Way—did not cause the problems, quality tools and techniques can be used to uncover and fix the root causes.
Leaders determine the strategy of an organization and focus everyone on accomplishing it. The lesson for leaders is to be careful what you focus on, because you may get more than you bargained for from your followers.
Michael P. Levy
Cause and effect fallacies
In the April 2010 article "Under Scrutiny," Mark Paradies writes that cause and effect analysis has "major shortcomings." One of these, he says, is that people jump to conclusions and only look for information that confirms their theory.
This can be true, but it is not how root cause analysis should be conducted, and it is not how I was taught to do it. The same argument can be made about any problem-solving or research effort, including the new cause and effect analysis he proposes.
He also writes that people look for one root cause, missing the possibility of other causes. That is not how I was taught, nor is it how I train people to do it. In Guide to Quality Control, Kaoru Ishikawa specifically instructed practitioners to be "certain all the items that may be causing dispersion are included."
Usually, there are several root causes to any problem. What Paradies failed to mention is the use of Pareto analysis, which allows the major causes to be addressed and the main impact of the problem to be resolved. For more than 20 years, people have been solving many problems and improving quality this way, even though they didn’t find all the root causes.
What is more disturbing is that Paradies writes, "Human performance issues (human errors) cause most quality problems." I once heard W. Edwards Deming say only 6% of the quality problems are caused by frontline workers. Deming said the workers are "handicapped by the system, and the system belongs to management."
Because managers are human, I suppose it still fits that errors are caused by humans. But Deming said one of the main obstacles to improving quality is to blame the workers, and problem-solving efforts should be focused on processes and systems, not the people. To attribute all quality problems to human error moves the focus away from the processes.
Paradies tries to prove the misconceptions of cause and effect analyses, and suggests some improvements. In my view, these improvements have always been part of effective cause and effect analysis. I guess it depends on how you’re taught.
In the article "Under Scrutiny," Paradies makes a good point regarding the overly simplistic single causal chain application of the five whys technique, but he does so by implying that problems occur without cause, and then goes on to posit that the root cause is the absence of a best practice and knowledge. This seems like another simplistic single cause to me.
Dismissing a useful technique because it may be misused seems ill advised. It makes sense to ask the whys for a variety of potential causation paths and use fishbone diagramming, which is based on the notion there are many potential root causes.
In the article, Paradies references the fishbone diagram as one of the old methods, even though effective use of the fishbone approach does exactly what he suggests in his new method, which is far from anything new.
The point of problem resolution is—and has been for centuries—to understand the problem, identify potential causes, single out the actionable ones, plan, act and adjust.
Problems arise from causes and conditions. It is the problem solver’s job to find those causes and conditions, and come up with a solution that addresses them. The problem solver must not lose track of the reality that in complex systems, you can never be sure of the effect any action may have. Often, there is interplay among a variety of causes and conditions.
We also should be aware we cannot wait until all cause possibilities are identified. Many root causes are not actionable. Analysis overkill is just as debilitating as jumping to false conclusions.