Abstract
Process capability indices have been successfully app lied to measure proc ess performance. Most oftraditional approaches for eva luating process capability are assumed that the underlying data are precise numbers, but it is much more rea listic in general to consider observed values are imprecise numbers. Thus, this study takes into consideration of a certain degree of imprecision on the sample data and develops fuzzy estimators for two one-sided capability indices Cpu and CPL by adopting the concept of Buck ley 's approach. In addition, a triangular shaped fuzzy membership function is constructed for testing process performance. Finally, this concept is also illustrated with a numerical example.