A0273
Title: Robust monitoring of process dispersion
Authors: Saddam Akber Abbasi - Qatar University (Qatar) [presenting]
Maha Amouna - Qatar University (Qatar)
Abstract: Control charts act as the most important tool of the Statistical Process Control (SPC) tool-kit for monitoring process parameters such as location and dispersion. Monitoring process dispersion plays a vital role in improving the quality and production of any manufacturing process. Control chart structures are mostly developed under certain distributional assumptions such as normality and no outliers. When these ideal assumptions are violated, the usual control charts can result in increased false alarms and low detection ability. We will present control chart structures that are based on robust estimation of process dispersion (interquartile range, median absolute deviation, Gini mean, etc.). The performance of these structures is evaluated for a variety of contaminated scenarios and also compared with the usual dispersion chart based on the sample standard deviation. A real-life example will be presented to illustrate the working of new control chart structures. The results will be helpful for quality control practitioners to choose a robust control chart for monitoring process dispersion.
