Setting and Using Product Specifications – An Overview

About this Article
This article by Mike Dobbert was presented at the NCSLI 2010 Symposium and published in the Proceedings. It won the "Dr. Allen V. Astin" Best Conference Paper award and also best Learning and Development award.

The version below is that published in the September issue of the NCSLI "Measure" journal.

• NCSLI Measure

Abstract

This paper presents techniques that manufacturers can use to set specifications and then describes how a metrologist can use those specifications in calibration. Specifications describe the warranted performance of a product or the expected accuracy of a measurement standard. From the manufacturer’s perspective, specifications must describe performance that can be achieved cost effectively. This paper looks at several statistical issues related to setting specifications. Techniques for characterizing expected product performance, considerations for drift and performance variation due to external environmental conditions, and the significance of measurement uncertainty are covered. A framework is presented that relates the expected performance of a product to its specifications while maintaining metrology goals, producibility, and competiveness. For calibration, specifications are often used as Type-B uncertainty estimators. Drawing from the statistical issues related to setting specifications, this paper discusses the use of product specifications in an uncertainty analysis.

View or Download

The entire paper is available to view or download in Portable Document Format. But if your browser doesn't already have the (free) viewer installed, you will first need to get and install the Acrobat® Reader™.

Setting and Using Product Specifications – An Overview