Sonu asks:

"Is zero defects possible? If so, what are the steps. We are dealing with around 300 parameters each having minimum 25 to 30 parameters to be met. We find it difficult to maintain zero defect for all parameters. Any thoughts???"

The first thing I would do is go to gemba. What does the machine, material, method and manpower (4M) for these processes look like? Then I would observe the process until a defect was created. It is always easier to catch the criminal in the act of committing the crime, rather than through detective work later. This in a nutshell is genchi genbutsu.

I would ask experienced people in the organization to share with me what they know about the problem, previous efforts to fix the problem, whether these were effective, and what is being tried currently to reduce defects. Their collective wisdom will inform and frame the issues.

These things would provide me with an intuitive sense of the products and processes.

The next step would be to review several types of Pareto charts of the types of defects, by part, parameter, by frequency and possibly other factors. The top items on the Pareto chart would be studied in further depth in terms of the 4Ms above, on the gemba, using a cause & effect diagram a.k.a. fishbone diagram.

At this point the combination of observed facts, intuition (what looks or sounds wrong, based on experience) and the data lead to the next steps. The next steps would most likely be a combination of things as simple as basic 5S (throwing out unnecessary items, putting all necessary items in the proper place for quick retrieval, and thoroughly cleaning to identify sources of contamination and filth), standardizing methods and procedures, checking whether our gages and measurement systems were capable, and possibly some design of experiment (DOE) type activity to see which parameters and conditions mattered the most.

I will leave the in-depth answer to this question of achieving zero defects to our friends Ron Pereira at the Lean Six Sigma Academy blog, Mike Wroblewski at Got Boondoggle?, and Rob Thompson at Quality Hero, each of whose Six Sigma chops far exceed my own.

Is zero defects possible? What is your experience? If you have sustained a zero defects process, what steps did you follow?