Nonparametric Statistical Process Control

2Basics of Statistical Process Control

Chapter Overview

In this chapter, we focus on the basic concepts, terminology, and some of the underlying statistical theory of univariate control charts. This includes topics such as control limits, placement of the control limits, sample size, sampling frequency and how to form rational subgroups. We start with the classical Shewhart control charts. Some of the more advanced, yet practical, topics such as the performance and the statistical design of a control chart are also introduced.

2.1 Basic Concepts

To describe and understand the underlying theory, we begin with a discussion of some important terminology.

2.1.1 Types of Variability

Within any process, a certain amount of natural variability, no matter how well the process was designed, developed, or maintained, will always be present. For example, not all car doors painted by the same machine under the same settings will be identical. Not all five‐pound bags of potatoes packed by the same plant will contain exactly five pounds of potatoes, no matter how carefully the bags are packed. Such variability is inherent to a process and cannot be (or need not be) totally eliminated, possibly not without incurring an enormous cost for a minimal return. Conversely, there can also be other kinds of variability present within a process that is disruptive, which can result in a deterioration of quality. The goal of statistical process control (SPC) is to identify the source(s) of this variability, ...