Quality Control
Simon Shaw s.c.shaw@maths.bath.ac.uk 2005/06 Semester II

1

Introduction

In quality control, we are concerned with problems involved in controlling the quality of a manufactured product. We’ll study two major techniques: 1. ACCEPTANCE SAMPLING where we are concerned with monitoring the quality of manufactured items supplied by the manufacturer to consumers in batches. The problem is to decide whether the batch should be accepted or rejected on the basis of a sample randomly drawn from the batch. 2. PROCESS CONTROL where goods are produced continuously and the problem is to detect changes in the performance of the manufacturing process and take action (when necessary) to control the process.

1.1

What is quality?

Most of us associate quality with luxury (such as a BMW car, a plasma screen television, . . . ) but by quality control we are thinking in terms of “things that work in the way we expect them to” i.e. “it does what it says on the tin”. 1. Quality implies fitness for use. 2. Quality means conformance to requirements. Thus, quality is defined by both customers and producers.

1.2

Variability compromises quality

Mass produced items are not identical. Some variation is inevitable and can cause problems. Too much variation might mean that parts which should fit together don’t. e.g. A screw might be too small/large to fit the corresponding bolt. There is a need to identify items which exhibit too much variation and deal with them, perhaps even scrap them.

1

1.3

Inspection versus Prevention

How should we deal with variation? An early strategy was to inspect goods at the end of the production line and release, to the consumer, only those of a sufficiently high standard. It was argued that the marginal cost of each item was small, so the cost of the rejected items was small. However, this argument ignored a number of