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

1.1 Early Pioneers of Industrial Engineering

Some have referred to industrial engineering as a “child of the Industrial Revolution.” Indeed, around the turn of the 20th century, the transition in the United States and elsewhere from an agricultural society to a society that relied on mass production of goods and services of all kinds provided the need for a new type of engineering. With labor specialization and new manufacturing technologies came the need for better methods for planning and organizing how people and machines interacted to form an efficient manufacturing system. In engineering schools this began in the early shop courses that taught students how new types of physical and chemical processes could transform raw materials into finished parts and products. As companies took advantage of new manufacturing technologies, jobs for technically trained people became readily available in many different hardware-oriented disciplines. In the early part of the 20th century, engineering schools soon found themselves inundated with students eager to learn how to efficiently produce goods and services for a variety of rapidly expanding markets. For example, around the beginning of the 20th century there were several hundred small shops producing different cars, or what was then called a “horseless carriage,” but by 1929, the US auto industry had consolidated many of these small shops into several very large corporations, including GM, Ford, Packard, Studebaker, Hudson, Dodge, and Chrysler. By carefully planning large production facilities, these companies were capable of producing more than 90 percent of the 32 million cars and trucks sold worldwide. And by 1930, this mass production capability resulted in there being one car for every five people living in the United States. The stories that follow are largely about how this type of transformation occurred and the role industrial engineering played.

The rapid industrialization, and its demand for engineers had a profound effect at the University of Michigan (UM). From 1880 to 1899 there were only four engineering degree programs at Michigan (civil, mining, mechanical, and electrical), and these granted a total of 530 undergraduate degrees and 30 graduate degrees over this19-year period. By 1950, the number of engineering programs at the UM had expanded to 22, and over a 12-year period from 1940 to 1952, these granted a total of 7,028 undergraduate engineering degrees and 2,176 graduate degrees.

As far as the practice of industrial engineering during the first few decades of the 20th century, it was mostly concerned with ways to improve worker productivity in various factories. This concern was facilitated by the writings of Fredrick Winslow Taylor and Frank and Lillian Gilbreth. In Taylor’s book The Principles of Scientific Management (published in 1911), he described a process of selecting the most productive worker in a group, documenting exactly how this person performed a job, and carefully timing that person’s movements. The combination of the careful documentation of worker motions along with the “time study,” as it became known, were then used as the basis for specifying standardized methods when performing jobs and for selecting and training other workers on how to perform the job in an efficient manner. The resulting motion times were referred to as “standard times,” which were provided to the managers and supervisors of a firm so they could better determine and manage labor costs, which often accounted for the majority of the total cost of a product. This was particularly true after Henry Ford shocked the industrial world in 1914 by offering to pay the best workers $5 per day (equivalent to about $120 today). This was twice the amount being paid to auto-assembly workers at the time, and it allowed Ford to hire the best mechanically skilled workers. This policy greatly reduced Ford’s 300 percent annual worker turnover rate and, thus, was shown to be a cost-neutral policy.

In his many writings Taylor advocated that the results from a time study could be used to test and select well-trained and highly motivated workers to perform a job. Several cases were provided in his 1911 book that showed very large increases in productivity by using the results of time studies of “first class” workers in various companies. Of course, the results of a time study also became the basis for piece/part pay systems wherein additional bonus pay was provided to workers who exceed the production goal.

In 1912, Frank Gilbreth wrote a primer on how to use Taylor’s principles. Gilbreth was an inventor, founder of a successful construction company, and later a management engineering consultant. He attributed a great deal of his early business success to Taylor’s principles of scientific management. One of the major insights Gilbreth articulated was that because “standard times” were derived from adding up the sequence of motion times demonstrated by skilled workers, it was also possible to determine how to reduce the total time required to complete an existing job by reconfiguring the required motions in a manner that contained fewer wasted motions. Gilbreth described a simple example of this principle: in his construction company pallets of bricks were sorted and moved by low-paid, unskilled workers to be closer to the skilled bricklayers constructing a brick wall. The pallets of bricks were also positioned by adjusting the scaffold on which the bricklayer stood to a height that did not require them to do much stooping or reaching. This repositioning of bricks and worker not only reduced the time required of the skilled bricklayers to complete a wall, since less time was now required to walk and carry bricks, but also reduced the fatigue caused by repeatedly stooping and carrying the bricks.

The value of Taylor’s scientific management principles became very well known and accepted after being used in a publically cited 1910 case before the Interstate Commerce Committee (ICC). In this action, the famous attorney Louis Brandeis solicited testimony from Taylor. This ICC forum allowed Taylor to publically describe his time and motion study methods and present his results from the time studies he performed for the railroads. The results swayed the ICC to deny the Eastern Railroad’s request for a ticket price increase. The ICC stated that the projected labor costs were far higher than necessary due to management’s inefficient use of workers.

Frederick Winslow Taylor (1900)

Together, Taylor and Gilbreth lectured between 1910 and 1915 at many different engineering and management conferences in the United States and abroad. Their common theme was that it was management’s responsibility to use quantitative techniques (e.g., time and motion study) to document and standardize how best to organize jobs and to select and train the best workers. They and their followers quickly became known as “efficiency experts” or “industrial engineers.” After the death of Frederick Taylor in 1915, Frank Gilbreth began to emphasize a slightly different philosophy than Taylor. He, with assistance from the Kodak Company, used the new medium of movies to film, sometimes at very high filming rates, skilled workers performing a variety of manual tasks. To do this, rather than time study workers in existing jobs, he established a laboratory where he could control the conditions in simulated work situations. In this sense he must be given credit as a pioneer for developing the first human motion capture laboratory in the United States. To acquire the time values that he needed he would place a clock with a large dial close to a worker and within the camera’s field of view. In this way he could time their specific motions very precisely by reviewing the films. Gilbreth then categorized the motions into fundamental, or “elemental” motions (e.g., reach, grasp, move, position, walk, carry). To each of these he computed an average time based on his empirical laboratory studies. By summing the individual motion times required for a person to perform a particular job, he asserted that he could predict the mean total time it would take any skilled worker to perform a job. This motion time prediction method quickly became a powerful industrial engineering tool. By using the Gilbreth time prediction method a trained job analyst could simply describe a job on paper by listing Gilbreth’s elemental motions. Since each of the listed elemental motions had an assigned time value, an analyst could now predict the total time for a person to perform a proposed new job. The Gilbreth motion-based time prediction methodology soon became part of the process used by engineers when designing new manufacturing operations. It provided industrial engineers with an estimate of the number and types of workers needed in a proposed factory, even before it was operating. It also provided a detailed standard description of the new jobs, which then improved the training of future workers.

Frank B. Gilbreth (1910)

After the much publicized Eastern Railroad case with the ICC in 1910, many industries took note of the writings and teachings of Taylor and Gilbreth. During the ’20s and ’30s these companies established motion analysis and time study departments staffed by industrial engineers. Improving the efficiency of worker performance, and thus reducing the cost of labor became a major goal of business executives during this era. Unfortunately, such a narrow focus had its limitations. In essence, workers were now being treated as a commodity in some factories. Supervisors would sometimes use time-study results to speed up a production line without regard to the mental and physical fatigue that it caused for some workers. Unionization of workers during the ’30s in such plants was largely attributed to the unsafe and inhumane working conditions caused by the excessive profit motivation of managers. Unions, such as the United Auto Workers, which was formed in 1937, negotiated (sometimes after striking) a fair day’s pay. They also required fatigue allowance times to be added to the standard time values. These fatigue allowances were based on how much workers slowed their motions during a day of work. Frank and Lillian Gilbreth’s book Fatigue Study: The Elimination of Humanity’s Greatest Waste, a First Step in Motion Study (1916) provided ample examples of how to determine if workers were becoming fatigued and how to reduce the amount of fatigue by redesigning the work conditions, such as providing stools to alleviate standing all day, arm rests on chairs, adjustable-height work tables, and adequate rest periods.

After the death of Frank Gilbreth in June 1924, his wife, Lillian Gilbreth, became a major spokesperson for those concerned with the use of motion and time studies in industry. While raising 12 children (their children Frank B. Gilbreth Jr. and Ernestine Gilbreth Carey chronicled the family’s life in two books, Cheaper by the Dozen and Belles on their Toes, which were later made into successful movies), Lillian Gilbreth continued lecturing to industrial engineers from various companies. Lillian, who had earned a PhD in psychology from Brown University in 1915, emphasized in her many lectures and conference presentations that a rigorous, empirical approach was necessary to gather and analyze the complex human micro-motion data one needed to accurately predict the cost of labor and provide less fatiguing work conditions. She also developed and advocated the use of worker surveys to quantify worker attitudes and opinions, stating that valuable information could be gained from worker input to guide workplace improvements. In this sense she became an early spokesperson for more worker participation in decisions to improve work conditions. Her leadership became very important following the Great Depression of 1929. During the ’30s and ’40s she served on several different high-level government advisory boards in the Hoover and Roosevelt administrations, providing guidance on how to organize efficient and safe workplaces.

Lillian M. Gilbreth (1921)

For this pioneering work, in 1965 she was inducted into the National Academy of Engineering as its first woman member. In 1966, she received the Hoover Medal, an award given for “outstanding extra-career services by engineers to humanity.” She has often been referred to as the “mother of ergonomics.” Many different engineering awards have been named in her honor, including awards from the National Academy of Engineering, the Institute of Industrial Engineering, and the Society of Women Engineers, to name a few. She gave many lectures at various universities and was granted the title of professor of engineering at Purdue University in 1940. During her career she received 23 honorary degrees, the first of these from the University of Michigan in 1928. Lillian Gilbreth died in 1972 at the age of 93.

Based on the early adoption of time study and motion analysis methods by many different companies in the early part of the 20th century, the first industrial engineering undergraduate degree program was established in 1909 at Pennsylvania State University to further expand the availability of people to practice in this field. In addition to courses in accounting and shop planning and layout, the Penn State program offered a course on scientific management that used texts from Frederick Taylor and Frank and Lillian Gilbreth. How the Gilbreths influenced the development of industrial engineering at the University of Michigan is discussed later in this chapter.

1.2 Origins of Operations Research—The Use of Mathematical Modeling to Support Management Decision Making

This section provides a basis for understanding how operations research, that is, the use of mathematics to model and analyze a wide variety of systems and situations, has become an intellectual pillar, along with ergonomics and information systems, in the evolution of the IE Department.

In the late 1830s Charles Babbage investigated the cost of transporting and sorting mail, a project that in 1840 resulted in England’s universal and highly efficient penny post. Much later, in 1908, two Swedish mathematicians—C. Palm and A.K. Erlang—established probability equations by which congestion in telephone systems could be represented. The resulting probability distributions of waiting times could then be used when designing the system.

In 1915, Ford W. Harris published a paper in Factory: The Magazine of Management titled “How Many Parts to Make at Once,” in which he introduced what is now known as the “economic order quantity,” a formula that is a cornerstone of modern inventory management. It allows a manager to balance the cost of reordering against the cost of manufacturing and maintaining a large inventory.

These are just three examples of the early use of mathematical modeling applied to operational problems. Improving, extending, and adapting these insights and results eventually led to whole fields of study and application: logistics, queuing theory (or, in England, “waiting line” theory), and production planning. Other similar precursors to contemporary approaches to analyzing operational problems can be found in the deep history of applied statistics and probability, game theory, graph theory, decision analysis, linear algebra, and so on, all of which have their antecedents in work performed by mathematicians in the 19th and early 20th centuries.

However, overshadowing these particular examples of the use of mathematics to address operational problems were the challenges to military decision makers in determining how to effectively operate in the world of 20th-century global warfare. This need for critical mathematical analyses of a wide variety of military problems eventually led to a discipline known by many names, including operations research, operations analysis, management science, and, more recently, analytics.

An early example of the use of mathematical modeling in the military is found in the 1916 book Aircraft in Warfare, the Dawn of the Fourth Arm, by English engineer F. W. Lanchester, which is a mathematical analysis of air-to-air combat that set the stage for a formal understanding of military conflict and its logistical components. (The resulting Lanchester equations, representing the evolution over time of power relationships among competing forces, are still used, not only in military analysis but also as a critical aspect of studying predator-prey situations). In 1917, Lord Tiverton showed the advantages of concentrating aerial bombs on a single target (a strategy leading to British and American forces using up to 1,000 planes in a single raid during World War II). Shortly after this, a study by the US Naval Consulting Board (headed by Thomas A. Edison) showed that the best way for surface ships to evade submarine attacks was to zigzag rather than to sail in a straight line.

In the unsettling years just before (and then throughout) World War II, the military establishments of the United Kingdom and their allies, including the United States, knew that winning the war depended on addressing operational issues. These included determining the best defensive and offensive actions against German U-boat attacks (which had been the cause of heavy losses of desperately needed supplies), finding the optimal ratio of escorts to merchant ships in cross-Atlantic convoys, determining locations for placing the newly invented radar towers and antiaircraft installations, developing the means to reduce the number of antiaircraft rounds needed to shoot down an enemy aircraft, and finding the most effective setting of the trigger depth of aerial-delivered depth charges.

Teams of mathematicians, statisticians, physicists, psychologists, and engineers in Great Britain, and then later in the United States, were charged with developing solutions for these kinds of operational problems. The involvement of first-rate scientists (including seven Nobel laureates) from a variety of fields created an environment within which their results had to be taken seriously.

The result was the development of a new military science by the end of World War II, called “operational research” in Great Britain and “operations research” (OR) in the United States. In the United States, military OR groups were first stationed at Princeton, Columbia, the MIT Radiation Laboratory, and other locations. Two important requirements for the success of the embryonic science emerged during this period: the use of multidisciplinary teams and the need for direct and prolonged field contact for analysts. Both the Navy’s Operations Research Group (later called the Operations Evaluation Group, or OEG), which included around 70 analysts by the time the war ended in 1945, required their analysts to be connected to a fleet or wing “client” for six months every year, and thus were given direct access to decision makers at the highest levels. The Army Operations Research Group had similar requirements.

After the war, these teams provided the foundation for a postwar expansion of the tools and approaches of OR. Some evolved into organizations, such as the Rand Corporation (from the US Army’s Project RAND) and the Center for Naval Analysis (from the OEG), that were formed to provide continuing scientific support for the military analysis of a wide variety of operational issues. Many others joined academic science, engineering, and mathematics departments at various universities.

George E. Kimball

Philip M. Morse (Photo courtesy of the MIT Museum)

Further academic interest in OR was evident in 1951, when Philip Morse (eventual chair of the MIT physics department, the first director of Brookhaven National Laboratory, and the founder and first director of the MIT Computation Center), along with George Kimball (a distinguished quantum chemist at Columbia University), published a newly unclassified book (originally written in 1945) titled Methods of Operations Research. This is believed to be the first comprehensive publication identified with the term “operations research.” In this book they defined OR as “a scientific method of providing executive departments with a quantitative basis for decisions regarding the operations under their control.”

Because of the authors’ previous experiences, the various methods discussed in their seminal book were all directed toward a specific set of operations: those that involved challenges to the military arising from World War II. Other OR analysts, after leaving military or government service, found opportunities to apply their analytical and mathematical approaches to a host of nonmilitary problems. These new operations, instead of involving the sinking of submarines or determining bomb loads, were integral to production, logistics, manufacturing, extraction, scheduling, inventory control, distribution, competition, marketing, and so on. In the ’50s the tools of OR began to be sought by a large number of firms, which motivated the hiring of well-regarded OR faculty members in the late ’50s when the IE Department was formed at the University of Michigan, as discussed later in this chapter.

1.3 Early Development of Industrial Engineering at UM

The relationship between the Gilbreths and the University of Michigan began very early in the 20th century when UM professor of mechanical engineering Joseph A. Bursley met the Gilbreths at a 1910 conference on the topic of scientific management. After returning from this conference, Bursley began studying the topic in earnest, as he was convinced of its importance to future mechanical engineering students. He requested and was granted a two-year leave to consult with various companies, including the New England Butt Company, where the Gilbreths were instigating and studying the effectiveness of time and motion studies. He returned to UM and, in 1915, established a course in the Mechanical Engineering Department titled “Scientific Shop Management.” During World War I this course was expanded to two courses, which were then required as part of the training of all Army ordinance officers. These two courses were so successful that the Army used them as a pattern for similar instruction at other universities.

Joseph A. Bursley (1924) – BS, Mechanical Engineering, University of Michigan (1899)

In 1921, Bursley became the dean of students at UM. In the same year Charles Burton Gordy was hired as an assistant professor of mechanical engineering for the purpose of continuing to teach courses on the topics of time and motion study.

After a series of committee reports and discussions, a new degree program in mechanical and industrial engineering was established in 1924. This new program required five years of study and 173 credit hours. Not surprisingly, eight years later only 14 students had graduated, probably because this bachelor of science degree required five years, whereas students in other engineering programs received a BS degree after only four years of study. As a result, the five-year degree program was replaced in 1934 with the awarding of a bachelor’s degree in engineering (mechanical engineering) after four years and a master’s degree in industrial engineering after the fifth year.

It should be clear from the preceding sections that in the early half of the 20th century industrial engineers were quite involved in organizing workers’ required manual tasks in ways that allowed them to be highly productive. The need for efficient labor productivity was even more evident during World War II, when men and women needed to quickly learn new skills and perform very complex tasks in new war industries. The interest in designing products and work that were more compatible with human capabilities continued to increase after World War II. This was particularly true in Europe and Japan, where a large proportion of the population was impaired and required assistive devices and accommodations at work. The rapid development of new products during and after the war stimulated the formation of teams of engineers and life and behavioral scientists with the intent of better understanding a variety of human-hardware interface problems. In 1950 this resulted in a UK organization known as the Ergonomics Research Society, now the royally chartered Institute of Ergonomics and Human Factors. In the United States these multidisciplinary teams formed the Human Factors Society in 1957, which is now known as the Human Factors and Ergonomics Society. These organizations held annual conferences and workshops and facilitated the writing of human factors design guides and other documents and books, such as Ernest McCormick’s 1957 textbook Human Factors.

Charles Burton Gordy (1950) – BS and AM, University of Pennsylvania (1919)

But it wasn’t only about developing better ways to utilize the labor force. In 1948 the American Institute of Industrial Engineers (in 1981 “American” was dropped from the name) was formed to promote the improvement and use of a variety of labor-planning tools as well as statistical and OR methods to improve productivity and the quality of manufacturing and service industries in the United States. By 1952, the increasing interest in military and nonmilitary OR led to the formation of the Operations Research Society of America (ORSA) as a means of advancing and diffusing developments in the field. This recognition of OR as a new intellectual discipline encouraged the establishment of educational programs devoted to developing and using the approaches and methods of OR. Depending on the nature of the institution, the academic homes for OR were usually offshoots of mathematics, statistics, business, or engineering departments (particularly industrial engineering), the latter being the eventual location of the University of Michigan’s OR-related courses and research activities.By the mid-1960s, a wide variety of OR methodological approaches were developed, including linear programming (e.g., the simplex method for solving linear programming problems was developed in 1947 by George Dantzig, a decade after he received his MS in mathematics from UM), dynamic programming, queuing theory, game theory, network analysis, replacement and inventory policies, reliability analysis, machine maintenance, scheduling, Monte Carlo simulation, decision analysis, and stochastic modeling. These approaches were applied to a variety of problem areas in industry and in other arenas, including manufacturing, marketing, transportation, communications, public safety, construction, health care, medical systems, banking, military operations, entertainment, hospitality, financial engineering, logistics and supply chain management, and commercial aviation.

1.4 Establishment of the UM Industrial Engineering Department

With the development of a sound statistical basis for time and motion study methods, combined with the newer operations research methodologies and, eventually, digital computing technologies, industrial engineering was beginning to be recognized as an important discipline in engineering schools and colleges in the United States. The effect at UM was that in 1946 the regents approved a degree designation of bachelor of science in engineering (industrial-mechanical); from then on, students could receive a BS degree in industrial engineering after four years of study. This degree option became very popular shortly after the end of World War II, especially with veterans, who saw it as a means to acquire jobs in the rapidly changing and highly competitive manufacturing sector.

In 1950, Gordy asked representatives from the Engineering Council for Professional Development (ECPD) to visit the program. They suggested several changes in the curriculum, namely a reduction in the number of accounting courses from the business school and an increase in courses in engineering economics, wage incentive and job evaluation methods, production control, plant layout, legal aspects of engineering, and statistics for engineers (the latter being taught by the Math Department). After UM made these changes, ECPD accredited the program. As a result, in 1951 the Board of Regents took the following action:

One year later, 98 students were enrolled in the new industrial engineering degree program within the Department of Mechanical and Industrial Engineering. Gordy, whose title was changed to professor of industrial engineering, became the program chair and was joined by assistant professors Quentin Vines, an expert in production planning; Wilbert Steffy, an expert in applied statistics and capital budgeting; and Edward Page, an expert in manufacturing processes and plant layout.

Quentin Vines – MS-ME (’53) University of Michigan

Wilbert Steffy – M.E.-ME&IE (’50) University of Illinois

Wilbert Steffy described the instructional program (The Michigan Technic, 1957) as follows:

Edward L. Page – BS-IE&ME (’37) University of Michigan

It is interesting to note that in the early part of the ’50s IE faculty research was mostly oriented to applied problem solving or consulting work in manufacturing companies. One exception was that the Methods-Time Measurement (MTM) Association, which supported a series of laboratory studies in the IE program. From 1953 to 1957, David Raphael worked with Gordy to perform research on the consistency of various standard time values within the MTM Association time prediction system, one of the most popular international systems for time prediction. His work led to a number of monographs concerning various aspects of predetermined motion times. This work was continued by Barbara Goodman when Raphael left in 1957. Walton Hancock, who was beginning as an assistant professor, then became the project director in 1960. In 1961, Hancock hired James A. Foulke, an electrical engineer, when Goodman left. The MTM Association project continued until 1972 and provided support for several PhD students under the direction of Hancock and Foulke. The emphasis also changed from determining time motion values to understanding the constraints on human productivity. By the end of the ’60s, statistically determined human learning curves, error rates, operator selection, local muscle fatigue, aging effects, and complex decision times were being studied in the newly established Human Performance Research Laboratory, which had gained support from several companies as well as the MTM Association. In contrast to the MTM Association research, which was largely laboratory based, various companies funded students to perform field studies in their plants. It is worth noting in this context that in 1953 Robert Carson received UM’s first PhD degree in industrial engineering with a dissertation titled “Consistency in Rating Method and Speed of Industrial Operations by a Group of Time-Study Men with Similar Training.”

By 1955, it was clear to the faculty that students and companies had a strong interest in industrial engineering topics and principles. The American Institute of Industrial Engineering and the Operations Research Society of America were now publishing scholarly papers by faculty members from several universities. The original four UM faculty members (Gordy, Vines, Steffy, and Page), who were teaching in the production engineering program (Option A as described by Steffy) within the Mechanical and Industrial Engineering Department, had established a curriculum that was based on strong analytical problem-solving methods and an understanding of how workers could be expected to perform in various industries. In other words, the graduates from this program were now being provided with the type of knowledge required to solve important strategic and tactical problems in a variety of complex manufacturing and service operations. Based on this, at the October 1955 Regents meeting the following request to form a new Department of Industrial Engineering was approved as submitted by then dean George Granger Brown:

Also, in 1955, Wyeth Allen was hired as professor of industrial engineering, He had expertise in organization management, had run an engineering consulting firm, and had served as president of a large manufacturing company. In the fall of 1956 the new Department of Industrial Engineering started offering courses with Allen as its first chairman. Several faculty members were added. In 1956, James Gage, an expert in engineering economics, and Richard Berkeley, who specialized in work measurement and labor planning. were hired. In 1957. Clyde Johnson, a highly respected management consultant was hired to broaden the curriculum to include organizational management and to develop student project courses for undergraduates.

Wyeth Allen – Hon Dr Eng (1953)

James Gage – UMBS, Mechanical Engineering, University of Wisconsin (1947)

Richard Berkeley – BS, Mechanical Engineering and Industrial Engineering, University of Michigan (1931)

Clyde Johnson – AB, University of Michigan (1931)

To develop the operations engineering area, in 1956 Robert Thrall, then a professor of mathematics at UM, joined the IE Department. In the previous decade Thrall had applied methods of mathematics and OR to military problems while consulting with the US Army Strategy and Tactics Analysis Group and the National Defense Research Committee. He also collaborated with John Nash on the potential application of game theory to military decision making. In 1962, he was instrumental in organizing the first US Army–wide Operations Research Symposium at the US Army Research Office, Durham, along with Dean Wilson, Merrill Flood, and Herbert Galliher (at that time the associate director of the MIT Operations Research Center and soon-to-be faculty member in the IE Department at UM).

Robert Thrall – PhD, Math, University of Illinois (1937)

At UM, further development in OR occurred a decade later when, in 1970, in collaboration with W. Allen Spivey (of the UM Business school), Thrall wrote Linear Optimization, one of the first textbooks devoted to linear programming. It should be noted that in 1969, Thrall was elected the 16th president of the Institute of Management Sciences (TIMS). During his term as president he was instrumental in founding and funding Center for Operations Research and Econometrics at the University of Louvain in Belgium, now recognized as a world center in OR. Thrall left UM in 1969 when he was asked to chair the new Department of Mathematical Sciences at Rice University, which had just been split off from the Department of Mathematics. While at Rice his research emphasis was the creation and development of data envelopment analysis.

Another important person in the OR area joined the IE department in 1959; Merrill M. Flood, who had received his PhD in mathematics from Princeton University. During World War II, Flood worked at the RAND Corporation, where he and Melvin Dresher conceived of and first analyzed the “Prisoner’s Dilemma” model of game theory, and the “Hitchcock Transportation Problem” (finding the minimum cost distribution of a product being supplied by several sources to several locations of use). He is also credited with naming the discipline of linear programming. Flood came to the University of Michigan with three titles: professor of industrial engineering, associate director of the Engineering Research Institute, and head of the Willow Run Laboratories (where he continued the work he started in the 1940s on combat surveillance systems). From 1959 through 1967 he was also professor of mathematical biology in the Department of Psychiatry in the medical school and senior research mathematician in the Mental Health Research Institute—foreshadowing the use of OR methods to study medical and health operational issues in which the department would become a leader. Flood was a founding member (and second president) of TIMS and, while at Michigan, was elected president of the Operations Research Society of America (ORSA). He was awarded the George E. Kimball medal in 1983 for his work with both societies and for his revolutionary contributions to OR and management science.

Merrill Flood – PhD, Math, Princeton University (1935)

Assisting Thrall and Flood in teaching OR during the late ’50s was Richard C. Wilson, who was appointed an instructor in 1956 while completing his PhD in facility and production planning methods in the new IE department. In 1961 he was promoted to assistant professor and two years later to full professor. (See more about Richard Wilson in the next chapter.)

After IE became a separate department in 1955 and hired well-regarded and experienced faculty members, the effect on the curriculum was immediate. Teaching the Option B set of courses that emphasized manufacturing processes, described previously by Steffy, was largely left to the Mechanical Engineering Department. The Option A courses consequently became the IE department’s primary teaching responsibility in the late ’50s, with changes being made as new faculty members were added. By 1960, the BS degree in industrial engineering required the following set of courses, and more than 250 undergraduate students were enrolled in the program.

Though the focus of the department in its first few years was still on methods to improve manufacturing operations, expansion into other problem arenas had begun. For instance, Merrill Flood was providing new analytical solutions on how to optimize the routing of trucks, plan health care services, and perform product pricing in highly competitive markets. Clyde Johnson established an IE student project program to reduce the cost of hospital care. Johnson’s pioneering hospital project course continues to this day, and over the years several faculty members have emphasized health care areas for their research in the department, as described in chapter 6.

1.5 Changes in the College of Engineering after World War II—Effects on IE

One change that affected the new department in the 1950s was a major philosophical shift that had begun after World War II within the College of Engineering leadership. Up until this time the major focus was almost exclusively on undergraduate education, mainly because that is what tuition and the state and federal government were funding. In the latter regard, the Servicemen’s Readjustment Act of 1944, which became known as the GI Bill of Rights, provided tuition payments for veterans and assistance with living expenses and books while studying toward a degree. During 1947, veterans accounted for almost half of all college enrollees, and by July 1956 about 7.8 million veterans of the 16 million that had fought in World War II had taken advantage of this program, thus making a college education available to those other than the rich in the United States.

Indeed, undergraduate enrollment in the college went from 2,155 in 1940 to 3,273 by 1970. With the hiring of additional faculty members in the college, many of whom had PhD degrees, by the early ’50s there was a growing emphasis on graduate education and research as well. The result was that graduate enrollment increased from 258 students in 1940 to 948 in 1970. As the new IE Department hired additional faculty members in the late ’50s and better research facilities were provided, the IE graduate programs began to develop. By 1957, an MS-IE degree was offered, though it did not have specific course requirements. In fact, the 1960 Engineering Bulletin stated: “A candidate for this degree must have completed satisfactorily the undergraduate industrial engineering degree or its equivalent, and must complete in residence a minimum of thirty hours of recognized graduate work approved by the adviser. The course selections necessary for this degree are rather flexible, but it is expected that approximately twelve hours of course study will be in the industrial engineering area.”

The IE program graduation profile in the graph displays this trend, with the early emphasis on providing undergraduate degrees followed by the initiation of the graduate program when the new department began in 1955.

The emerging need to provide a strong graduate education in the late 1950s, which soon became a formal second function of the college, had a very important goal: to establish a first-class research university. Doing so was contingent, however, on the development of sponsored research because Michigan legislators were much more interested in providing well-educated engineers with BS degrees. For instance, federal and private funding of the college’s research was only $215,700 for the 1940–1941 academic year (equivalent to about $3.6 million today). But with the growing emphasis on research, and with the hiring of new faculty who had PhD degrees in the 1960s, there was the interest and capability to develop more sponsored research. Fortunately, following World War II many companies needed the results of engineering research to guide the development and use of new technologies and methods of conducting business. The federal government had also become accustomed to supporting university research during World War II, and this continued with the establishment of the National Science Foundation in May 1950. As a result, by the 1965–1966 academic year, sponsored research funding from all sources had reached about $10 million in the College (equivalent to $74 million today), which was more than double the state allocation for the entire general operating fund for the College. In essence, the College rather quickly became dependent on the faculty’s ability to attract major sponsored research funding. As will be discussed, the IOE faculty responded in kind to this policy.

Another College-wide change during the 1950s was the beginning of the North Campus development. The resulting new building space enabled the College to integrate science and research with graduate engineering education by establishing a variety of laboratories. The Cooley Laboratory, dedicated on October 24, 1953, was the first building on North Campus, and it provided space for some important developments in electronics, much of which had been done previously as classified, military-oriented research at the UM Willow Run Laboratory during World War II. Over the next few years, several other engineering buildings were completed, including the Ford Nuclear Reactor in the new Phoenix Laboratory in 1955. This facilitated a new educational program in nuclear engineering, one of the first in the United States. Other buildings quickly followed, including the Automotive Engineering Laboratory in 1957, the Aeronautical Engineering Laboratory and Wind Tunnel in 1957, and the Fluids Laboratory (later renamed the G. G. Brown Laboratory Building in honor of the dean who initiated the building construction on North Campus), which followed a 1956 plan that had been proposed by Dean Brown and approved by the regents. In this plan, all the functions of the College were to move to North Campus in phases. Unfortunately, it was not until the 1980s that such a move took place.

The delay in the construction of the additional buildings necessary to provide the space for all of engineering to be located on North Campus had a major effect on the IE program. Fortunately, the new IE program initiated in the mid-1950s required only a modest amount of dedicated laboratory space in West Engineering, mainly for the research sponsored by the MTM Association. Almost all of the other department space at that time was used for offices and classrooms. This was to change in the ’60s with the growing interest in solving a variety of human performance problems that were plaguing companies at the time.

West Engineering and the “Arch” were home to the IE Department from 1955 to 1983.

1.6 Synopsis of 1950s—IE Contributions and Major Events

• 1955: IE department separates from the Mechanical and Industrial Engineering Department. Four faculty members (Gordy, Vines, Steffy, and Page) from the Mechanical and Industrial Engineering Department become the founding members, with Richard Wilson as a lecturer.
• 1956: The undergraduate curriculum is established and the first IE courses are taught. Six new faculty are added to the department (Allen, Gage, Berkeley, Thrall, Flood, and Johnson).
• 1957: Johnson establishes a senior project course in hospital systems, IE undergraduate enrollment is now 260 students, and several OR courses are added to the curriculum at the undergraduate and graduate levels.
• 1958: Graduate-level courses are established in manufacturing process planning, plant layout, optimization methods in IE, hospital systems, and quality control and inspection. These courses enroll about 20 master’s-level students in Ann Arbor. The part-time evening master’s program enrolls 18 students in Flint. The first IE Department PhD student graduates (Fritz Harris), supervised by Allen.