The First 50 Years of the Department of Industrial and Operations Engineering at the University of Michigan: 1955–2005

2.1 General Trends in IE Education and Research—A Period of Growth

By 1960, the newly formed Industrial Engineering (IE) Department was certainly part of the larger trends in the College of Engineering. Not only had the undergraduate program grown during the late ’50s but in the ’60s the graduate programs became very important. This can be seen in the graph depicting the annual IE graduation rates during the ’60s. Some of the large increase in the number of master’s degrees awarded in the late ’60s was due to the fact some of the faculty were teaching night classes at the University of Michigan Flint Campus and at a community college in Saginaw, Michigan. About 20 MS students graduated annually during the late ’60s and early ’70s through this IE extension program. It was discontinued in 1982 due to low enrollment and lack of interest in teaching the off-campus courses in the evening. Some would say this was the beginning of the department’s withdrawal from actively promoting and supporting continuing education for professional engineers.

Providing the means to expand the PhD program in IE required funding for research, and several IE faculty members began to aggressively seek funding. In 1962, the total sponsored research funding in the department was about $25,000, but by the end of the ’60s more than $700,000 was acquired annually from various companies and federal agencies (the equivalent of about $4.2 million today). This funding allowed the more research-oriented faculty to “buy out” from teaching a course or two during the academic school year and supported additional PhD students, a trend that continued through the ’70s and ’80s. Such fund-raising work was further necessitated by the diminishing university support for the College during the ’70s, which will be discussed later.

2.2 The Need for a Broad IE Discipline

During the early ’60s many technological changes were occurring in the world. The race to understand outer space was in full swing, with satellites flying overhead and the launching of spacecraft that flew astronauts to the moon and back as part of the NASA Apollo Program. The federal government continued to develop very sophisticated military equipment, including high-altitude spy planes to help determine what the rest of the industrialized world was doing. Most importantly, programmable digital computers were rapidly proliferating in universities and companies. These allowed complex computational problems to be solved in hours that had taken days, or worse, could not be solved at all without the use of these new computers. It clearly was an exciting time for science and technology. In many cases, new research results were completely changing the work, play, and lifestyles of most people living in the industrialized world, and IE education in the United States was rapidly expanding, with 48 degree-granting departments in operation by the end of the ’60s.

2.3 The Importance of the Computing Environment in the Department

Prior to the widespread use of digital computers in academia, special digital processors were built to solve specific types of engineering problems. In 1952, the Michigan Digital Automatic Computer (MIDAC), built by the University of Michigan (UM) Willow Run Research Center, was the first digital computer at the university. Several MIDACs were built and used for solving linear-equation problems in the 1950s, including transportation problem algorithms developed by Professor Merrill Flood and facility layout algorithms of interest to Professor Richard Wilson.

Richard Wilson using a UM-built MIDAC in 1961 in West Engineering.

In the ’50s and early ’60s, more flexible, programmable digital computers were quickly spreading to businesses and universities. Unfortunately, programming such machines was still time consuming and often required instructions and data to be input on hundreds of key-punch cards. These cards were then submitted to a computer center that had the machines to read the hole patterns in the cards and run the program on a mainframe computer. The results would be returned to the programmer as a set of printouts.

In 1952, the university acquired a card-based computer, the IBM Card Programmable Computer. In 1956, the first generally available computer, an IBM 650 was acquired and set up on Central Campus in a new computer center that would serve all faculty and students. This was followed by a succession of larger and more powerful computers over the next 10 years, but all required students to bring their punched cards to the center and wait, sometimes hours, for the staff to run their programs. As computer interest and programming capabilities exploded on campus, two developments relevant to IE took place. First, in 1965 a time-sharing capability was provided to users with remotely located teletype-style input/output terminals being connected to the mainframe computer, an IBM 360. A few of these terminals were located in West Engineering, so IE students and faculty could now type in their instructions and data rather than trudging to the computer center with a box full of punch cards. Unfortunately, the fees to use the remote terminals were about three times higher than that for running a program in a batch mode at the computer center.

The second development was that computer companies like Digital Equipment Corporation and Hewlett-Packard began to build small, stand-alone, scientific computers. Such a machine was acquired by the IE department in 1963 and used by Wilson and others to develop algorithms for solving transportation, production scheduling, plant layout, and other problems. By 1969 these mini computers were being used to monitor experiments online, and some were interfaced with the mainframe computer for later data processing. For example, an HP2115A mini computer was installed in the G. G. Brown Laboratory Building and used by the IE human performance laboratory and the nuclear engineering, bioengineering, and mechanical engineering departments. In addition, James Foulke (a very creative electrical engineer) worked with Professor Walton Hancock to create a special, programmable, digital data calculator that was configured to collect and analyze human motion and time data and was interfaced with mini computers for further data analysis.

Richard Wilson using the department’s new Litton General Purpose LGP30 Digital Computer (with a drum memory) in 1963 to run plant layout and transportation models in West Engineering.

Many years later, in the mid-1980s, the College established a sophisticated network known as the Computer Aided Engineering Network (CAEN) to support various types of computers, including the new personal computers. Now all faculty, staff, and students had access to computers to solve many different types of engineering problems. This was further aided by the College administration’s negotiation with Apple Corporation to provide up to 60 percent discounts for the Macintosh personal computers. From then on most large engineering problem solving was accomplished with the aid of analytical models and associated algorithms that could be run on individual and network-supported personal computers.

The HP2115A computer used to monitor real-time experiments in 1969 on North Campus.

The data calculator used for MTM Association research in 1969 in the G. G. Brown Laboratory Building on North Campus.

2.4 Changes in IE—More Research, More Graduate Students

The rapid changes in manufacturing, transportation, construction, health care, and other industries in the 1960s meant engineering students had to be cognizant not just of a specific type of new technology but also how that technology functioned in unison with other existing technologies commonly used in a particular industry. Managers were asking questions such as the following: (1) What costs would be incurred when adopting a new technology? (2) How will a proposed technology be used to enhance the manufacturing, operation, and maintenance of new products and services? and (3) How could a variety of people benefit from using a new technology? Answers to these and other related questions led to the hiring of additional faculty members in the department. These faculty members expanded the breadth of IE courses, which then provided a broad technical education at the undergraduate level and a certain amount of specialization at the graduate level, the latter relying on a defined set of optional courses or “tracks” in various subspecialty areas in which the faculty had particular expertise.

2.5 Faculty Changes

Teaching courses to address the types of questions posed earlier at both the undergraduate and graduate level required the hiring of additional faculty members. These new faculty members were highly committed to conducting their own research while teaching courses and advising PhD students. The department had six full-time, tenure-track professors by the end of the 1950s, 10 by 1962, and 18 by 1969. Of the faculty members who joined the department during the 1960s, the following had a major and lasting effect.

Walton M. Hancock – PhD, Johns Hopkins (1954)

Walton Hancock was hired in 1960. His research stressed the understanding of how people perform and learn complex perceptual-motor tasks with emphasis on statistical modeling of human performance in manufacturing systems. He soon established a human performance laboratory that broadened time and motion study to include statistical models of human performance variations caused by different types of training, age, and fatigue factors. One of his many original contributions was the learning curve, a statistical method for forecasting how quickly workers could be expected to learn new skills. In 1963 he was promoted to department chair, a position he held until 1968. During the 1970s he used his analytical modeling methods to improve patient scheduling and hospital staffing processes and held a joint appointment in the Department of Hospital Administration within the School of Public Health. Later, Hancock served as associate dean of the College of Engineering and director of the Center for Research on Integrated Manufacturing from 1985 to 1989. (See Hancock’s memoir at the UM Faculty History Project: www.um2017.org/faculty-history/faculty/walton-m-hancock.)

Richard C. Wilson – PhD, University of Michigan (1961)

Richard Wilson, who had begun his academic career at UM as an instructor in 1956, was promoted to professor in 1961. As an instructor in the late ’50s he had taught the use of digital computers for solving various IE problems. He was soon acknowledged for his expertise in providing useful solutions to complex analytical models related to facility layout, production planning, and production scheduling. In the mid-1960s he established a production simulation laboratory for teaching and research related to facility layout and production scheduling. He later served as associate dean in the College of Engineering from 1968 to 1972 and chair of the department from 1973 to 1977. In 1978, he established a collaborative education and research program in manufacturing with mechanical engineering and computer science that later became the Center for Research on Integrated Manufacturing.

Herbert Galliher – PhD, Yale (1952)

Herbert Galliher was hired in 1963 from MIT, where he had served as the associate director of the pioneering Operations Research Center for seven years. When he came to Michigan, Galliher taught and developed courses in inventory and production analysis, stochastic processes, and mathematical modeling. As one of the founders of the field of operations research, his research and teaching covered a range of applications. He developed and named the “stuttering Poisson” distribution to describe demand processes in supply situations. He was one of the first to use queuing theory to analyze landing congestion of aircraft and to use stochastic models to represent disease processes, leading to an early and influential body of literature on the management of cervical cancer, arteriosclerosis, and breast cancer. The results of his project, Aircraft Engines: Demand Forecasting and Inventory Redistribution, as well as the development of a military depot system simulator revolutionized the forecasting of the demand for spare engines and associated logistical decisions. Galliher was the founding editor of International Abstracts in Operations Research and associate editor of Operations Research and the SIMA Journal on Applied Mathematics. He was also a well-respected consultant to a variety of corporations and government agencies. (See Galliher’s obituary in the December 11, 2000, issue of the University Record: www.ur.umich.edu/0001/Dec11_00/22.htm.)

Bertram Herzog – PhD, Engineering Mechanics, University of Michigan (1961)

In 1965, Bertram Herzog transferred to the IE Department from the Department of Engineering Mechanics, where he had become an expert in the use of computer graphics for the design of complex structures. In 1968, shortly after joining the IE faculty, he became the first director of the new Michigan Educational Research Information Triad (MERIT) computer network, one of the first operational computing networks in the country. Under his direction, MERIT connected the mainframe computers at the three large public universities in Michigan in 1972. This enabled faster computational processing than had been possible and the rapid sharing of large data files. By the mid-1980s, the MERIT staff started managing the connectivity of various computers around the country as part of the NSFNET, the precursor of today’s Internet. (See Herzog’s obituary on the Electrical Engineering and Computer Science Department website: http://www.eecs.umich.edu/eecs/about/articles/2008/Herzog.html.)

Don Chaffin – PhD, University of Michigan (1967)

Don Chaffin, who had been an instructor in the department while studying for his PhD (1964–1967), returned in 1969, after serving as a faculty member at the University of Kansas Medical Center. Working with Hancock, he enlarged the scope of the Human Performance Laboratory to include the study and prevention of working conditions that raised the risk of physical fatigue and injury to a worker’s musculoskeletal system. From 1977 to 1981, he served as department chair. In 1982, he assumed directorship of the Center for Ergonomics (discussed in chapter 6), remaining in this position until 1998. From 1998 until his retirement in 2007 he directed the Human Motion Simulation Laboratory within the Center for Ergonomics. Chaffin is a fellow in eight scientific and professional societies, was elected to the National Academy of Engineering in 1994, and received the National Engineering Award from the Association of Combined Engineering Societies in 2008. (See Chaffin’s personal page on the UM website: www-personal.umich.edu/~dchaffin/.)

Katta Murty – PhD, University of California–Berkeley (1968)

Katta Murty joined the department in 1968 from the University of California, Berkeley, replacing Robert Thrall, who had taken a position as chair of the Applied Mathematics Department at Clemson University. Like Thrall, Murty had a deep understanding of and commitment to furthering the development of linear and nonlinear optimization methods and algorithms. His eight textbooks have been used as the basis for teaching optimization methods and applications to students around the world. (See his personal page on the UM website: www-personal.umich.edu/~murty/.)

Daniel Teichroew – PhD, University of North Carolina (1953)

Daniel Teichroew joined the department in 1968 from Case Western Reserve University. Before that he was a professor at Stanford University. By the time he came to UM, he had already published six books on management science with an emphasis on computer algorithms and languages. Teichroew was hired as the department chair, a position he held until 1973. During this period he continued his pioneering development of a Problem Statement Language (PSL), which was the intellectual core activity within the Information System Design and Optimization System (ISDOS) consortium that he had begun building while at Case Western Reserve. The resulting ISDOS software and technical reports provided an efficient way for a person to use a computer to completely specify the information technology (IT) requirements for a proposed engineering or management system. In essence, the ISDOS reports provided a set of coding rules that would guide the formal statement of a technical or managerial problem in such a manner that a computer algorithm, known as a Problem Statement Analyzer (PSA) could evaluate the PSL problem statements and thus determine if the IT requirements were met. (For more on ISDOS, see chapter 6. See also Teichroew’s memoir at the UM Faculty History Project: http://um2017.org/faculty-history/faculty/daniel-teichroew/memoir.)

Ralph Disney – DEng, Johns Hopkins (1964)

Ralph Disney came to the university in 1962 as a visiting professor from the University of Buffalo (now the State University of New York at Buffalo). He rose quickly to full professor at UM after completing his DEng degree in 1964 from the Johns Hopkins University. His adviser at Johns Hopkins was a well-known expert in applied probabilism, Bruce Clarke, who soon after joined the UM Mathematics Department. Disney and Clarke continued their collaboration at UM and authored two textbooks on applied probability and stochastic processes in the early ’70s. His research contributions were in queuing theory (particularly queuing networks such as those found in manufacturing, transportation, and communications) and the mathematics of stochastic processes. He produced more than 70 research articles and was a founder of the Applied Probability College of Operations Research Society of America. In 1977, Disney accepted an invitation to be the Charles O. Gordon Professor of Industrial Engineering at the Virginia Polytechnic Institute and State University in Blacksburg. Disney was recognized by the Institute of Industrial Engineers (IIE) with the David Baker Award and the Albert Holzman Award for teaching, research, and other contributions to his field as well as the Frank and Lillian Gilbreth Award, the highest honor bestowed on an industrial engineer. In 1999, Disney was elected into the National Academy of Engineering. (See Disney’s obituary from the November 15, 2014, Washington Post here: http://www.legacy.com/obituaries/washingtonpost/obituary-print.aspx?n=ralph-l-disney&pid=173179805.)

Seth Bonder – PhD, Ohio State University (1965)

Seth Bonder joined the department in 1965, after receiving his PhD in industrial engineering from the Ohio State University. From 1952 to 1956, he served as a pilot for the US Air Force in the Korean War. From this experience and his education, Bonder became a major leader in the use of operations research methods to solve very large and complex military problems of all sorts. In 1972, he left the department to establish and become chief executive officer of one of the most successful military OR consulting companies in the United States. The company, Vector Research Incorporated (VRI) grew to more than 400 employees, including several PhD graduates from the department. His work garnered him many national awards, and he was elected into the National Academy of Engineering in 2000. Before his death in 2011, he shifted some of his considerable talent to directing VRI programs in health care. He also endowed a scholarship program in the IOE Department. (See the INFORMS website for a tribute to Bonder written by Stephen Pollock: https://www.informs.org/ORMS-Today/Public-Articles/December-Volume-38-Number-6/In-Memoriam-Seth-Bonder-1932-2011.)

Stephen Pollock – PhD, MIT (1964)

Stephen Pollock joined the department in 1969. Before that he was on the faculty of the US Naval Postgraduate School. His PhD from MIT provided him with a deep understanding and background in a variety of empirical and probabilistic-based operations research methodologies. He used these insights to collaborate with many other faculty members in solving problems related to (1) public systems, for example, political redistricting, nuclear arms, and collusion monitoring; (2) medical systems, for example, diagnostic methods, staffing levels, and early infection detection; (3) manufacturing, for example, supply routing for ship building, motor vehicle emission testing, and spot welding classification; and (4) sporting events, for example, swimmer event assignments and golf handicapping. From 1981 to 1990 he served as department chair. His stochastic operations modeling work in a variety of complex problems gained him a number of national awards, and he was elected to the National Academy of Engineering. (See his page at the College of Engineering: http://pollock.engin.umich.edu/.)

Richard Jelinek – PhD, University of Michigan (1965)

Richard Jelinek joined the department in 1965 after completing his PhD on health care planning methods. Working with Clyde Johnson, he developed several well-funded research studies that demonstrated how a variety of industrial engineering methods could be used to improve the quality and cost-effectiveness of local hospitals. In 1970, he cofounded the Medicus Company to apply the results of these studies in a wide variety of hospitals, and in 1971, he joined the faculty in the UM School of Public Health, Bureau of Hospital Administration. Jelinek later became chief executive officer of AmeriChoice Corporation, a national health care consulting company. (See his executive profile on the BloombergBusiness website: http://www.bloomberg.com/research/stocks/private/person.asp?personId=228637&privcapId=24792&previousCapId=33254122&previousTitle=
RedBrick%20Health%20Corporation.)

Among the other faculty members who joined the department for short periods during the 1950s and 1960s were Edgar Sibley (1966–1972), who was hired as an associate professor and worked with Daniel Teichroew, and Dean H. Wilson (1961–1968), an associate professor hired to assist in teaching the use of digital computers to control manufacturing systems and to develop and teach the use of Monte Carlo simulation methods. Assistant professors included Fred Black (1955–1958), Richard Evans (1958–1963), Hugh Bradley (1965–1968), Stephen Kimbleton (1968–1973), and Richard Baum (1969–1976). Richard Pew, an assistant professor in psychology, had a joint appointment in the department from 1966 to 1972.

On average there were 17 faculty members in the department during the 1960s, and these provided a certain amount of structure and stability in the curriculum; prerequisite courses were offered more often than before, and more advanced courses were provided to support the growing graduate program. But toward the late 1960s, the end of the Apollo space program, the contentious war in Vietnam, and greater demands for racial equality in various organizations (including higher education) led society to lose interest in technology, and the University of Michigan executives began a process of shifting financial support away from the College of Engineering and toward the liberal arts, social, and health-related programs. This defunding of the College, which continued throughout the ’70s, meant restrictions on hiring new faculty, limited salary increases, and difficulties in achieving promotion to tenured positions for the existing new faculty members, especially if the candidate did not have a large amount of sponsored research funding to support his work and students.

In regards to this latter situation, even though the number of full-time faculty members had increased from nine to 18 in the ’60s, the number of funded positions on the general operating budget supported by the state and tuition funds only increased by one during the same period. In other words, there was an enormous burden on the faculty to acquire grant and contract funds to support themselves and their graduate students. In fact, the college administration expected all tenured faculty members to provide at least 20 percent of their academic-year salary from external funds. The IE faculty for the most part responded by ramping up the externally funded research from about $25,000 in 1962 to more than $700,000 in 1970 (the equivalent of about $4.2 million in 2014). This relatively large amount of sponsored research in the department supported at least 43 PhD students by the end of the ’60s and allowed the faculty to grow to 19 members at the beginning of 1970 without placing a major burden on the diminished college general fund budget.

2.6 Beginning to Move to North Campus

The increasing sponsored research funding throughout the ’60s also allowed some of the empirically oriented IE faculty members to establish a few laboratories for instruction and research purposes in the G. G. Brown Laboratory Building on North Campus. About 2,000 square feet was dedicated for these IE laboratories and offices by the mid-1960s. This relatively small space accommodated about 15 PhD students and five faculty members for the purpose of physically simulating and studying production and assembly operations as well as conducting a variety of human performance experiments.

Unfortunately, the very tight college budgets provided by the state starting in the mid-1960s and continuing through the 1970s resulted in almost no funds being available to construct the buildings needed to move more of the College operations to North Campus. As a result, faculty and students in many departments, including IE, had to commute daily between the instructional spaces in West and East Engineering Buildings on Central Campus and their research offices and laboratories on North Campus. This situation was finally resolved when the entire IE faculty moved to North Campus in 1983 (see chapter 4).

The west end of the G. G. Brown Laboratory Building in 1964 is shown in the foreground; this is where IE offices and laboratories were located on the North Campus.

2.7 1970 IE Curriculum

By the end of the 1960s, the undergraduate IE curriculum continued to emphasize a balance between traditional, hardware-oriented engineering technologies that dealt with materials, structures, chemicals, electricity, and electronics and the more analytical and statistical methods needed to solve multifactor problems in highly complex systems composed of people operating and servicing hardware and software systems of all kinds. Candidates for the degree Bachelor of Science in Engineering (Industrial Engineering) had to complete the following program:

By the end of the ’60s, there continued to be no specific course requirements for the MS or PhD degrees. The MS degree required candidates to take a total of 30 credit hours (selected from 400- or 500-level courses), 18 hours of which had to be industrial engineering courses. Students entering the master’s program had to have completed a BS in industrial engineering or the equivalent. Students who did not have a BS in industrial engineering could be required to take additional courses in statistics, linear programming, or computer programming without graduate credit.

The graduate courses in the department (described later) were clustered into the following categories, each of which had at least two faculty members:

1. Human performance (Hancock and Chaffin)
2. Management engineering (Gage, Johnson, Jelinek, and Steffy)
3. Operations engineering and management science (Thrall, Murty, Disney, Baum, Kimbleton, Galliher. and Bonder)
4. Process systems design (R. Wilson and R. Evans)
5. Computers and information processing systems (Teichroew, Herzog, D. Wilson, Merten, and Sibley)

The PhD students were required to take at least two graduate courses from three of these areas and one graduate course from each of the other two areas. They then had to pass both a written and oral examination of their knowledge in the field. This test was referred to as a steering review and was normally taken at the end of the candidate’s second term in residence. Students were also required to demonstrate a reading competency (either by testing or course work) in two foreign languages by the time they graduated, one of which could be satisfied by taking nine credit hours in a cognate area approved by the adviser. With successful completion of the steering review, the PhD students were asked to select members for their doctoral committee and to present a written and oral proposal to them, which represented their thesis preliminary examination. If the proposed thesis was accepted, the student would commence the research and, at some point, present the results in both a written dissertation and a public final thesis defense presentation before the doctoral committee.

2.8 Synopsis of the 1960s—IE Contributions and Major Activities

• IE faculty size doubles from nine to 18 full-time members.
• Sponsored research funding increases from $25,000 in 1962 to $700,000 in 1969.
• A popular off-campus IE master’s degree program grows with classes offered in the evening at Flint and Saginaw sites.
• The PhD program increases from about 10 students to more than 40 students.
• The health care systems area, under the direction of Clyde Johnson and Richard Jelinek continues to expand with student projects in several area hospitals.
• The Human Performance Laboratory, under the direction of Walton Hancock, expands its motion and time prediction research to include models of human fatigue and operator learning.
• Space in the North Campus G. G. Brown Laboratory Building is acquired for research offices and laboratories.
• The Defense Systems Laboratory, under the direction of Seth Bonder, completes a number of studies modeling military operations at all levels.