Applying Six Sigma statistical analysis for process improvement

Problem

At a manufacturing client, Non-conformance Reports (NCRs) were being generated when they received materials that were out of specification limits. Large amounts of key engineering, quality, and operations resources were being used resolve and implement the disposition of these out-of-spec materials. The results was that it diverted these critical resources from value-added work. Management needed to get control of the sources of the problems and identify potential solutions.

Approach

The process was mapped and the relevant data was extracted from a vast QA database. Standard Six Sigma statistical analysis tools were used to identify useful patterns in performance for those parts of the process that were owned and controlled by the client.

Results

Problematic product categories and failure modes were prioritized and the Subject Matter Expert (SME) resources required for their resolution were quantified. The data clearly showed that subset of the SME pool carried the bulk of the load and that two of the SMEs clearly had mastery over best practices. By using best practices and leveling, the load across all SMEs could reduce the lead time for NCR resolution to 10–20% of historic levels.