Where Does Quality Begin?
It begins with the PDCA cycle and an understanding of variation--not with ISO 9001
by Dale K. Gordon
Every day quality professionals struggle with the eternal question that keeps them employed: How do we make things better?
At a recent meeting of an organization's quality professionals, I asked, "What do you believe the sources of your troubles are?"
They all had detailed explanations for individual problems, but my question was about the quality system. They sat silent for a minute and then agreed they had a quality system certified by an accredited third party and approved by their customers. "How could the cause of our problems be the quality system?" they asked.
I then asked what was probably the most difficult question of all: "Can you describe what your quality system is and how it prevents the troubles you are having?" This, of course, was greeted with more silence and finally agreement that their quality system was ISO 9001.
And therein lies the problem.
Many people--managers in particular--define the quality system of an enterprise as ISO 9001 or any identified standard their system is built around. ISO 9000:2000 defines a system as a "set of interrelated or interacting elements" (clause 3.2.1) and defines a quality management system as "a management system to direct and control an organization with regard to quality" (3.2.3).
If we go back to W. Edwards Deming's views, we are reminded of his views of a "system of profound knowledge," which consisted of:
- Appreciation for a system.
- Knowledge about variation.
- Theory of knowledge.
Deming said the various segments of the system of profound knowledge cannot be separated but interact with each other. Thus, knowledge of psychology is incomplete without knowledge of variation.
The same can be said of any quality management system. It is a collection of subsystems, processes, knowledge, variation and people who execute the system, all interrelated and managed by those charged with that responsibility. It is not ISO 9001 or any other standard. The quality system is whatever the organization decides it is with all its defined (or undefined) processes, people, variation and knowledge inherent in the system.
In the same meeting of quality professionals, I next asked this obvious question: "If you related all of your problems to a paragraph of the standard on which the quality system is built, then added all the findings from internal and external audits against your system, and then performed a Pareto analysis of the nonconformances found in the system, what would you conclude is the problem with your quality system?"
The responses came back almost instantaneously: corrective and preventive action followed by process control and supplier control. "These are related to more than 80% of the problems we are having," they answered.
I thought about this for a moment and then asked about design control or, in the new vernacular, planning of product realization. The heads shook mightily: "No problems there; in fact, we hardly ever have any findings in that area."
I contend thinking that way is part of the problem.
Shortchanging "plan" and "act"
We tend to be real good at the "D" and "C" (do and check) parts of the PDCA cycle, but plan and act always seem to get shortchanged. In ISO 9001:1994, clauses 4.4.1 (design planning) and 4.4.4 (design inputs) indicate the design process must consider customer, regulatory and statutory requirements.
The same holds true for the new version (ISO 9001:2000) under paragraph 7.1 (planning of product realization) and 7.3.2 (design and development inputs). Both indicate the need for structured planning and consideration of specification requirements and customer, statutory and regulatory needs.
Nowhere in these sections is there anything about learning from process and product failures and then designing or redesigning to eliminate them. The ISO 9001:2000 revision even goes so far as to include a process/continuous improvement cycle description and reference to the PDCA cycle. But it does not mention that the design stage is where the planning is most important and where the acting needs to take place.
So we ask ourselves some of the following questions:
- What is it about our quality systems that does not function correctly?
- Why do preventive and corrective actions always seem to top the list?
- Is it because we lack the fortitude to make the structural changes in our systems that would truly prevent problems from recurring?
- Is the investment in fixing that much less than the investment in preventing the problem to begin with?
- Do we really design for manufacture and use what we learn from our process control failures to redesign the product?
- Do we really listen to our suppliers and understand why they fail to interpret our design descriptions or process specifications and repeatedly fail to meet specifications?
- Do we work with the manufacturing organization and suppliers to match process capabilities to design needs?
- Do we really understand the sources of variation in our processes (both mechanical and human) and have actions in place to eliminate them or at least eliminate the ones that have the greatest impact on our ability to meet requirements?
More and more, the answers to these questions are coming from programs or initiatives such as Six Sigma, lean thinking or the old value engineering or reengineering all the way through the supply chain.
Many times these activities are seen as being outside and separate from the quality system or as extra management efforts aimed at improving both customer satisfaction and cost control. The issue with these programs or activities is they are applied well after the product is designed and the failures have occurred.
So there are more questions:
- How do we measure the design activity?
- When the design activity is audited, how knowledgeable are the auditors about the design processes and their relationships to the manufacturing and customer delivery processes?
- How well do activities such as design reviews take into account the ability of the procurement process to translate design ideas into tangible product?
These questions are not found in the ISO 9000 or spin-off quality system requirements. In fact, when you look for discussion of corrective action in ISO 9001:1994, you find there is little that points back to the origin of the product. Clause 14 (in ISO 9001:1994) refers more to the process of corrective action and to the documentation of the process than to an evaluation of the causes all the way back to the original design assumptions. It focuses on systems rather than individual processes.
The same is true of ISO 9001:2000. While the combination of clauses 8.4 and 8.5 does a better job of indicating the action is more important than the documentation, the standard is still focused on the generic corrective action process and its execution rather than on the thorough examination of causes all the way back to the design element.
Rarely do I find Six Sigma, kaizen, value engineering or other similar activities are the documented preventive action part of a quality system.
Response of automotive and aerospace industries
Automotive and aerospace are among the industries that have addressed this quality system problem in a strict ISO 9000 world by adding requirements. In the automotive world you find such additions as advanced product quality planning (APQP), potential failure mode effects analysis (FMEA) and production part approval process (PPAP), in addition to the ISO 9001 based QS-9000 requirements.
In aerospace you find FMEA, control of key characteristics (AS9103) and strict process change control, in addition to AS9100, another ISO 9001 based requirement.
Are these requirements considered part of the quality system? Sometimes they are not, because they are not strictly defined as ISO 9000 requirements. But they are necessary parts of the system and need to be treated as such.
So where does that leave quality professionals? The lesson here is twofold:
1. ISO 9001 or other standards are not the quality system. The system is a culmination of whatever the organization says it is through the processes, activities and knowledge it uses to meet product specifications and customer needs. The documentation and execution may meet ISO 9001, but I suspect much that is a part of an organization's quality system is not included in the quality system documentation.
2. The system is only as good as its design and execution. Its design includes the product design. If we do not design the system so we can learn and incorporate that learning into new processes, designs and systems, we are doomed to repeat the mistakes of the past. Then we'll go chase some new thing (such as Six Sigma) while deriding our quality systems, people or processes for being inadequate for the modern age.
So I take you back to Deming and Walter M. Shewhart (actual creator of the PDCA cycle) and tell you variation is the enemy and the quality system is more of an internal culture than an actual thing. You have to understand how all the elements of the system fit together if you want to make an improvement.
The PDCA cycle does work, when accompanied by knowledgeable management and some good luck.
DALE K. GORDON is director of quality and business improvement for Rolls-Royce Corp.'s Defense North American Business Unit in Indianapolis. He is chair of the American Aerospace Quality Group and was one of the writers of the AS9100 standard. Gordon earned a bachelor's degree in industrial engineering from General Motors Institute (now Kettering University) in Flint, MI, and a master's degree in business administration from Butler University in Indianapolis.