In today's complex high-tech manufacturing landscape, with myriad smaller suppliers and multicomponent systems, establishing robust supply chains is crucial for large (high-volume) customers. We require tight tolerances on LSL and USL. Smaller suppliers are usually single-source, and given the small volumes requested and limited resources, our specification requirements’ negotiating power is limited. 

A JMP-driven process capability decision flowchart addresses these challenges. By utilizing the normal approximation approach, we can effectively assess process capability, ensuring we consider all data points to avoid missing critical information. By using normal approximation (e.g. Distribution (Fit All)/Life Distribution), we find the best (GoF) curve fit to identify and address normality violation modes to pursue process improvement.

If Ppk < 1.33, we determine whether it's a spread (Pp) or a k-shift problem. For k-problems, we conduct a one-sample t-test against the target and if it fails, equivalence testing of the mean at specified thresholds (margin and alpha). For Pp problem, we conduct one-sample equal variance tests against spec range and. if it fails, equivalence testing of standard deviation. 

Equivalence tests (Fit Oneway) incorporate negotiating thresholds beyond standard hypothesis tests accounting for business models and quality risk agreements between suppliers and customers. If the Pp or k-shift requirement is not met to address the problem, we collaborate to conduct process improvements then reestablish quality requirements. By using Workflow Builder to systematize our flowchart, we ensure consistent decision making and establish mutual trust.

Our aim is joint profitability and growth, borne from a transparent, systematic, and continuously improving problem-solving approach.