Gabel, Stanley H. (1990, ASQC) Eastman Kodak Company, Rochester, NY
When implementing SPC within a manufacturing or service area, one quite often has to contend with many different processes, each with its own set of "key process variables" which need to be controlled or monitored. Such a scenario can place a large burden, in terms of data management, summarization, reporting, and most of all interpretation on the personnel in charge of managing this process. The PROCESS PERFORMANCE CHART (PPC) is a graphical tool which facilitates this process and provides a presentation technique which clearly and quickly highlights any potential problem areas.
The PPC provides in a graphical format many of the quality measures that are frequently used to evaluate the performance processes. The basic approach is as follows: Many of the commonly used measures, such as PCI (Process Capability Index), are merely functions of the process' performance in terms of its mean and/or standard deviation, often stated relative to the specifications of the process. The PPC is simply a plot of the variability (standard deviation) of a process vs its-deviation from aim (mean-aim). In order to plot many process variables with varying scales of measurement in a single graph, it is necessary to normalize these values to a single scale. The normalizing factor used by the PPC is the specification range for the variable. This normalization has added benefits: it allows us to map regions of constant PCI values, constant CPK values (another commonly used capability index), and even constant PPM (Parts Per Million) levels. Now at a single glance one can evaluate which process variables are operating at an unacceptable mean level or at an unacceptably high level of variability, and determine their associated PCI and CPK values. With this charting mechanism both measures are available in a visual display, along with the basic process measures which drive them.