BEST SELLER - Process Capability Analysis
William A. Levinson, P.E., FASQ, CFPIM is the principal of Levinson Productivity Systems, P.C. He is an ASQ Certified Quality Engineer, Quality Auditor, Quality Manager, Reliability Engineer, and Six Sigma Black Belt, and the author of several books on quality, productivity, and management.
Attendees will learn how to calculate process capability and process performance indices that are often required by internal or external customers of manufacturing processes. The webinar will show the relationship between a Six Sigma process and its performance index. The webinar will then go beyond this textbook material to cover processes that do not follow the normal (bell curve) distribution, and for which the traditional calculation methods do not work properly. An off the shelf and generally accepted method will then be presented for these non-normal applications.
Process capability analysis is a key activity in quality management. Process capability and process performance indices reflect the ability
of a process to meet customer requirements, and are therefore often
required by internal and external customers. Process capability analysis
ties in with statistical process control because (1) a successful
process capability analysis requires the process to be in a state of
control and (2) estimation of the parameters, traditionally the process
mean and process standard deviation, is required for both activities.
- Know the difference between process capability, which reflects only short-term variation as measured by the sample range or sample standard deviation, and process performance, which reflects all variation sources. If the rational subgroup (sample that reflects all variation sources) has been selected properly, the two indices should be roughly equal.
- Relate the process performance index to the nonconforming fraction or defects per million opportunities (DPMO).
- Test the distributional assumption (traditionally the assumption that the process data conform to the normal or bell curve distribution) before relying on the capability and performance indices, and select another distribution if the original assumptions are not met. This also carries over into statistical process control, where reliance on the normality assumption for a non-normal distribution can generate excessive false alarms (out of control signals).
- Calculate meaningful performance indices for non-normal process distributions.
Who Should Attend
Manufacturing managers, quality professionals, and purchasing
professionals; the latter with regard to the need to determine whether
capability reports from suppliers make sense. That is, has the supplier
relied on the normality assumption without first checking it?
- Webinar Code: BEST-WILE-0005