Preamble for special issue honouring John F. MacGregor

Authors


This special issue of the Canadian Journal of Chemical Engineering is dedicated to Professor John F. MacGregor, in celebration of his 65th birthday.

John grew up in St. Catherines, Ontario. In high school John MacGregor, John Grace (Canada Research Chair at UBC) and David Dodge (former Governor of the Bank of Canada) were classmates. John MacGregor completed his undergraduate studies in chemical engineering at McMaster University in Hamilton, Ontario in 1965. He went to the University of Wisconsin-Madison, where he obtained an M.S. in Statistics and an M.S. in Chemical Engineering in 1967. John studied with William G. Hunter, an extraordinarily accomplished statistician and chemical engineer. From 1969 until 1972 John worked for the Monsanto Company in Texas City. He returned to the University of Wisconsin to obtain his Ph.D. in statistics in 1972. John's thesis advisor was George Box, an eminent statistician. John started his academic career in the department of chemical engineering at McMaster University in 1972. He served as department chair for the period 1988-1991. He held the Dofasco Professor of Process Automation and Information Technology and is one of very few to hold the title of University Professor.

Over the past 35 years, John has made significant contributions to research and development at the interface of chemical engineering and statistics. John has worked in system identification, process control, adaptive control, statistical quality control, polymer reaction engineering, and multivariable statistics. In the area of statistical process control, John wrote a paper in 1990 in the Journal of Quality Technology that challenged one of the fundamental assumptions of statistical process control espoused by Edward Deming, who is considered the originator of total quality management. The assertion was that feedback control, or ‘tampering’ would increase process variability. Like most of John's papers, his approach was to clearly articulate the problem that Deming had addressed and to demonstrate its limitations in the environment in which most chemical processes operate. It is an understatement to state that this paper generated controversy in the quality control community!

The development of multivariable statistical methods has been the focus of John's research activities for the past 20 years. He has pioneered the development and application of these methodologies for process monitoring and quality improvement for steady state and batch processes and new product development. The objective in multivariate methods is to make more efficient use of the vast quantities of process and product data that are often available. More recently, he has been using image analysis to solve process engineering problems. His 1991 publication, co-authored with J. Kresta and T.E. Marlin, ‘Multivariate Statistical Monitoring of Process Operating Performance,’ Canadian J Chemical Engineering, 68(1), has over 250 citations and is one of John's most highly-cited papers.

John's research output is prodigious. He has published over 200 referred journal publications. A close colleague, Dr. Tom Marlin, described his contributions as follows: “… they are targeted to solve large classes of problems not addressed by existing methods. He more than masters the fundamentals, he combines keen insights into the real physical problems to develop solutions whose methods and assumptions match the real situations, not convenience for publishing papers”.

John has been particularly interested in seeing academic research adopted by industry and has been instrumental in the success of two academic-industrial consortia. He was founding Co-Director of the McMaster Advanced Control Consortium (MACC) and the Associate Director of the McMaster Institute for Polymer Production Technology. Companies have had early exposure to emerging research, they have been able to share experiences, faculty members have benefited from industrial perspectives and graduate students have been provided with tremendous opportunities to interact with companies. John has worked on applications of multivariate methods in mineral processing, food processing, groundwater and soil data, texture analysis of products, forest and steel products, combustion and is now branching out into exciting applications in medical diagnostics.

John has mentored a large number of graduate students. Nearly 60 master's students and 40 Ph.D. students have had the privilege, myself included, of working with John. In 1996 he received the Presidents' Award for Excellence in Graduate Student Supervision from McMaster University. His former graduate students have gone on to make their own significant and independent contributions in the field of process systems engineering in academic and industrial positions.

John has also taught many influential industrial short courses, such as Advanced Process Control (with P. Taylor), SPC Interfaces (with J. Stuart Hunter and me), and Multivariate Statistics (with S. Wold). Through these courses, he has reached thousands of practicing engineers and statisticians.

John has been the recipient of many awards and honours. In 2007 he was named a Fellow of Royal Society of Canada (FRSC). He is a Fellow of the Canadian Academy of Engineering and of the American Statistical Association. His industrial outreach was recognized in 2003, when he and Dr. T. Kourti, along with industrial partners Dofasco and Tembec received the Natural Sciences and Engineering Council (NSERC) Synergy award. He is recipient of the W.G. Hunter Award and the Shewhart Medal, both from the American Society for Quality Control, the Herman Wold Medal from Swedish Chemical Society, the Professional Engineers of Ontario Engineering Medal (Research and Development), the Bell Canada Forum Award and the Century of Achievement Award from the Canadian Society for Chemical Engineering.

John continues to contribute to research as an Emeritus Professor at McMaster and to applications through a spin-off company that offers software and services for applications of multivariate statistics.

As a former student of John's, I have greatly benefited from his insight and in particular, his approaches to problem definition and solution.

As guest editor of this issue, I would like to express my appreciation to the authors and referees. I would also like to thank Sandra Peake-Thibodeau of the Journal office for her assistance and advice.

Ancillary