Detection of Process Control Shifts Using Shewhart (X-Bar) and Cumulative Sum (CUSUM) Charts

Abstract

Abstract: This study compares how the Shewhart (X-Bar) chart detects small, difficult-to-identify shifts and variations in process control, as the Cumulative Sum (CUSUM) chart does. The production process of bread in a Bakery in Gboko, Benue State, Nigeria was studied for four months (August, September, October and November). Data were collected on the daily weights (kg) of shortening fat systems in every mixture in four shifts daily; the Shewhart (X-Bar) chart was used to ascertain whether  the process was within or out-of-control, i.e., whether the quantity of shortening fat fused into the production process daily was adequate or not for the bread produced. It was discovered that the production process most consistently met quality standards for shortening weights in November, followed by August, then September, and least effectively in October; based on the number of out-of-control days that the CUSUM chart was able to capture, which the Shewart chart was unable to capture. It was concluded that the CUSUM chart indeed facilitates the detection of small variations that are not typically identified by the X-bar chart.