Soren Bisgaard’s Contributions to Quality Engineering

Description

Søren Bisgaard was an extremely productive and insightful scholar of modern industrial statistics and quality engineering. He was amazing for both his breadth of interests and the depth of his scholarship. Søren was one of the very few people making substantial contributions in so many basic areas in statistics and quality engineering.
This compilation collects 31 of his works and is divided into four broad areas:
Design and Analysis of Experiments
Time Series Analysis
The Quality Profession
Healthcare Engineering
This book provides a comprehensive coverage of essential statistical methods for the 2k-p factorial system and shows the basic principles of time series analysis through examples. Furthermore, this book presents the connection between the application of the scientific method and quality improvement, and it points out the importance of quality improvement to tangible financial results. Finally, this book explains the seemingly paradoxical idea that we can enhance quality while reducing cost of healthcare.

Sujets

• Business
• Qualité, hygiène, sécurité, environnement
• Sciences & Techniques
• Statistiques
• Statistiques
• Techniques
• Automatisation et Contrôle
• Contrôle des processus
• Automatisation
• Automate
• Quality Control
• Sciences et techniques
• Søren
• Technology
• Engineering
• Probabilités et statistiques

Informations

• Publié par: ASQ Quality Press
• Date de parution: 27 avril 2017
• Nombre de lectures: 0
• EAN13: 9781953079367
• Langue: English
• Poids de l'ouvrage: 1 Mo

Extrait

Søren Bisgaard’s Contributions to Quality Engineering
Ronald J.M.M. Does, Roger W. Hoerl, Murat Kulahci, and G. Geoffrey Vining Editors
ASQ Quality Press
Milwaukee, Wisconsin
American Society for Quality, Quality Press, Milwaukee 53203
© 2017 by ASQ
All rights reserved.
Library of Congress Cataloging-in-Publication Data
Names: Bisgaard, Søren, 1951-2009, author. | Does, R. J. M. M., editor.
Title: Søren Bisgaard’s contributions to quality engineering / Ronald J.M.M.
Does, Roger W. Hoerl, Murat Kulahci, and G. Geoffrey Vining, editors.
Other titles: Contributions to quality engineering
Description: Milwaukee, Wisconsin : ASQ Quality Press, [2017] | Includes
bibliographical references and index.
Identifiers: LCCN 2017012321 | ISBN 9780873899567 (hardcover : alk. paper)
Subjects: LCSH: Quality control. | Experimental design. | Time-series
analysis. | Medical statistics. | Bisgaard, Søren, 1951–2009.
Classification: LCC TS156 .B566 2017 | DDC 658.5/62—dc23
LC record available at https:/lccn.loc.gov/2017012321
No part of this book may be reproduced in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher.
Director of Products, Quality Press and Programs: Ray Zielke
Managing Editor: Paul Daniel O’Mara
Sr. Creative Services Specialist: Randy L. Benson
ASQ Mission: The American Society for Quality advances individual, organizational, and community excellence worldwide through learning, quality improvement, and knowledge exchange.
Attention Bookstores, Wholesalers, Schools, and Corporations: ASQ Quality Press books, video, audio, and software are available at quantity discounts with bulk purchases for business, educational, or instructional use. For information, please contact ASQ Quality Press at 800-248-1946, or write to ASQ Quality Press, P.O. Box 3005, Milwaukee, WI 53201–3005.
To place orders or to request a free copy of the ASQ Quality Press Publications Catalog, visit our website at http://www.asq.org/quality-press .

Preface: An Introduction to Søren Bisgaard’s Body of Work
Søren Bisgaard was an extremely productive and insightful scholar of modern industrial statistics and quality engineering. Unfortunately Søren passed away in December, 2009 at the age of 58. Many of us felt that the best way to honor his memory was to compile a selection of his published works into this volume. Søren was very proud of his affiliation with ASQ and a large proportion of his works appeared in ASQ journals. It was only natural that we would approach ASQ’s Quality Press to publish this work.
Søren’s total opus was much too large and too rich for a single volume, even if we restricted our attention to those works that appeared in ASQ journals. Hence, a major challenge that we faced was selecting the specific manuscripts included in this volume. We all struggled with the final decision on which specific papers to include. To put things into proper perspective, four times Søren won ASQ’s Brumbaugh Award that annually goes to the paper published in an ASQ journal that makes the greatest contribution to the field of quality control.
Søren was a true visionary, which made some of these decisions very difficult. Many of his papers are relatively timeless. Others were important as preludes to other, more foundational work. Some were ahead of their time, for example, “The Future of Quality Technology: From a Manufacturing to a Knowledge Economy and from Defects to Innovation,” for which Søren posthumously won the Brumbaugh Award in in 2013. This paper was Søren’s Youden address at the 2005 Fall Technical Conference.
We divided Søren’s works into four broad areas: Design and Analysis of Experiments Time Series Analysis The Quality Profession Healthcare Engineering
Each editor selected what he considered the most important manuscripts and ordered them according to broad themes. Søren was truly amazing for both his breadth of interests and the depth of his scholarship. Søren was one of the very few people of making substantial contributions in so many basic areas in statistics and quality engineering.
Those of us who knew Søren well miss our colleague. With the passage of time, we see more and more how important he was to our profession. We also realize more and more how we miss him, the person, especially his laugh, his love of good food, and his love of good conversation. Of course, most of all, we miss our friend.
Part I
Søren Bisgaard’s Work on the Design and Analysis of Experiments
G. Geoffrey Vining
I associate Søren most with his work on the design and analysis of experiments; it also happens to be my primary research area. I remember many late-night conversations with Søren at professional conferences where we talked about the proper approaches for planning and analyzing experiments.
We very quickly discovered that we were the next generation of scholars within the George Box school of thought in a time becoming more and more dominated by “optimal” design enthusiasts. We gladly accepted and embraced that responsibility. This perspective shaped my paper selections.
The first paper in this section is “Must a Process Be in Statistical Control Before Conducting Designed Experiments?,” winner of a Brumbaugh Award. I consider this paper Søren’s best contribution to this area. It is a very accessible paper, well written for the quality engineering practitioner. It has virtually no mathematics. In fact, it is an excellent description of Fisher’s uniformity trials. In the process, Søren shows his training by Box and the resulting influence of Fisher through Box.
Many industrial statistics consultants reflect a rather naïve understanding of proper design and analysis of experiments, especially those who come out of the Deming school of quality improvement. Their single most important tool (in some cases, truly their only tool!) is the Shewhart control chart. These consultants, who have trained many engineers over time, insist that no one can conduct an experiment unless the underlying process is in control. Søren, very effectively, efficiently, and patiently, shows the contrary. This paper is a must read for all serious quality engineers and industrial statisticians.
Søren’s and my shared research area is response surface methodology (RSM). The seminal paper for this area is Box and Wilson (1951). RSM is a sequential learning strategy that uses a sequence of small experiments to discover “good” operating conditions for a product or process. Many people use optimization techniques to specify these conditions; however, they rarely are truly optimal, but they are very good and much better than the initial conditions. RSM experiments range from very small Resolution III two-level fractional factorial designs to very large second-order experiments.
Søren’s RSM passion was the 2 k–p factorial system. He was extremely well trained at the University of Wisconsin in how to generate these designs and their alias structures, and how to fold them over to fit additional terms. He was very good at the basic abstract algebra required to master these systems.
The next five papers in this section reflect Søren’s love of the 2 k–p factorial system. We present these five papers in chronological order, which allows us to see some of the development in Søren’s thought. Each demonstrates Søren’s ability to communicate even abstract concepts extremely clearly. Each paper illustrates these concepts through carefully selected examples.
The first paper, “A Method for Identifying Defining Contrasts for 2 k–p Experiments,” is from 1993, when the Taguchi parameter design was popular. Søren addresses some very important issues with regard to the contrasts required to generate appropriate small 2 k–p designs. The Taguchi approach often involves the study of a relatively large number of factors, which influences Søren’s choice of examples. Søren shows the basic principles for selecting the defining contrasts. The examples provide excellent insights.
The second paper in the set of five is “Blocking Generators for Small 2 k–p Designs.” Blocking is an extremely important aspect of the proper planning of experiments. Very small 2 k–p designs compound the problems presented by blocking. Søren has three nice examples to illustrate his basic points. However, the section “An Unexpected Benefit of Blocking” is the real jewel. He discusses a real experiment, including the thought processes of the team. He explicitly considers the “instability of the external process conditions” (a major theme in the first paper of this set). He later points out the unexpected benefit: If something goes wrong in one block, only it needs to be repeated. More importantly, blocking takes care of any time effect as the result of repeating the block. Thus, the blocking procedure provides a “damage control” mechanism.
The third paper, “A Note on the Definition of Resolution for Blocked 2 k–p Designs,” picks up on a theme presented in passing in the previous paper. This paper is rather short, but it demonstrates Søren’s scholarly thoroughness: What does design resolution mean when we have blocks? He again effectively explains the subtlety that blocks present to a proper interpretation of design resolution. He clearly demonstrates how to derive the appropriate alias structures for several designs in order to illustrate his proposed revised definition of resolution.
The fourth paper in this set of five, “The Design and Analysis of 2 k–p × s Prototype Experiments,” is not only important in its own right but also important as a precursor to the last paper in this set. Søren begins to look at split-plot experiments, where some of the factors are hard to change a