The first chance event: meeting a girl. In 1954, I was a sophomore studying mechanical engineering with an aeronautical option at Carnegie Mellon University (in those days called Carnegie Institute of Technology) and was enjoying a Christmas vacation at home in Harrisburg, Pa., with my old crowd of high-school friends. One of the girls in the crowd had a visitor from upstate New York who was a freshman at Syracuse University. We found ourselves mutually attracted and started dating during that and other vacations, interspersed with letters (alas, no e-mail then) and phone calls. Things continued in a fairly serious vein.

Meanwhile, back in college, after a recruiting visit in my junior year to the Glenn L. Martin company in Baltimore (they actually flew a plane up to Pittsburgh to take a bunch of us budding aero engineers to Baltimore to show off their facilities), I became somewhat disenchanted with working as an aeronautical engineer. This was partially due to being shown one particular room where 300 engineers sat, each in front of a desk calculator (alas no desktop or laptop computers in those days). So I decided that perhaps graduate school in some sort of management curricula might be an option rather than making my fortune in industry right away. Since I was interested in being near my sweetie, one of the schools I applied to was Cornell. The closest thing I thought to management and still being in engineering was industrial engineering, and that's the program to which I applied.

By the time it was necessary to choose a graduate school, my sweetie was no longer sweet — we had broken up early in my senior year. As it turned out, the best financial offer I had regarding scholarships and assistantships was from Cornell's Department of Industrial and Engineering Administration (now the School of Operations Research and Information Engineering, nee School of Operations Research and Industrial Engineering, nee probably a few other name changes from its original). I was admitted to the master's of industrial engineering (MIE) program, which was a terminal two-year master's degree for those going into industry. I took the offer.

The next chance event: meeting a professor. When I arrived at Cornell in the fall of 1956, the department, under the able leadership of Andrew Schultz Jr., was just discovering a relatively new area relating to industrial engineering called operations research. Non-military O.R. was just emerging, and its original World War II military uses continued to be expanded. Remember, this was only four years after the founding of the Operations Research Society of America (ORSA) and three years after the founding of The Institute for Management Sciences (TIMS).

I took a course in the introduction to operations research, where we spent most of the time trying to define O.R. (not much has changed in the ensuing 50 years, has it?). But I also had to take undergraduate statistics, since in those days it was not required for undergraduate engineers, and I had no idea what a standard deviation was (in engineering we used safety factors rather than designing for two or three sigma events). So, as chance would have it, I signed up for course 3541, Introduction to Statistics, taught by Professor Robert E. Bechhofer. Not only was he a great teacher, but one of the most nurturing and caring individuals I have ever met.

I did very well in my first statistics course — I got an A, which drew Professor Bechhofer's attention. He was always on the lookout for good statistics students to persuade to continue their studies. He called me into his office one day and suggested that I change to the academic M.S. program, write a thesis and consider going on for the Ph.D. Bob Bechhofer was very low-key but very persuasive. I figured it would cost me maybe one extra semester if I stayed during the summer to write a master's thesis. And I thought, gee, being addressed as doctor would really make my parents happy (although that wasn't quite the kind of doctor they had in mind). So I changed to the M.S. degree program and spent the summer of 1958 writing a master's thesis. And this produced another chance event.

Meeting the right girl. By this time, I had decided not only to complete the master's thesis but to go on for the Ph.D. as well. Early in the summer, a mixer was held for summer students and for those of us staying on campus during the summer. I had just come out of the French exam — in those days, for a Ph.D. in any field, one needed to pass two language exams (aren't you current O.R. students envious?). I headed to the mixer, since the sister of the chap I was rooming with at the time, I'll call her Jeannie (because that was her name) was one of the summer students, and I had noticed this gorgeous summer student blonde around campus who I was sure would be at the mixer and to whom I wanted to get introduced. By the time I found Jeannie, the blonde was ringed three deep with guys. So Jeannie said, "Look, I'll introduce you to her later; let me introduce you to this other nice girl in my dorm." The "nice" girl wasn't blonde, but she was nevertheless a pretty cute brunette, so I said, "Sure." Well, it turned out that the blonde, later that summer, ran off with an Ithaca bus driver, and the cute brunette and I got married a little more than a year after that mixer. Again, hooray for chance (and I still say this after 48 years, and not just because that cute brunette, who is now a bit gray but still very cute, is looking over my shoulder).

Chance again. Having decided to pursue the doctorate, I had to eventually choose the area of the field in which to work (it was starting to bifurcate into optimization and stochastic process). We had to take a balanced program of courses, so I took courses in linear programming, inventory control and queueing theory. One of my professors, Lionel Weiss, had a large grant to study stochastic inventory control and said he would support me if I wished to do a dissertation in that area. Lionel became my thesis advisor and essentially said anything I wanted to do I could, as long as it fit his grant parameters.

So, again, chance as well as dollars took me into a thesis in multi-stage stochastic inventory control (about which, at the time, I knew very little). But I loved applied probability and especially the queueing theory course I took, so I was to make my career in chance, as well as by chance.

It was now 1960, and the digital computer was making its appearance on the scene. Cornell had just gotten an IBM 650, an addressable drum machine. We all had to take a course in computer programming (which at the time was assembly language — no FORTRAN yet.) After spending many evenings in the computer room where the 650 was housed — debugging programs by looking at the array of lights on its consol — I decided that there was no way I was going to do a thesis that depended on computing, although the department was getting interested in Monte Carlo simulation techniques. Ironically, years later, I did dabble a bit in discrete-event simulation research and teaching, but in my dissertation writing days, simulation packages (e.g., ARENA) weren't even a gleam in anyone's eyes. In fact, I don't think most of the package inventors were even born yet. So I embarked on a rather theoretical thesis, involving stochastic inventory control with transshipping.

A couple of non-chance events. Through the efforts of Professor Schultz, the first two summers after arriving at Cornell, I obtained summer jobs with the Atlantic Refining Company (now ARCO) in Philadelphia. My first assignment was working on an LP model of gasoline blending, using a desk calculator applying the simplex algorithm term-by-term to a very large matrix (and you wonder why I decided to pursue a dissertation in the stochastic area rather than optimization?). After finishing up my dissertation, I decided to make my fortune and go into industry, accepting a full-time position with Atlantic.

As I had been in ROTC at Carnegie Tech and graduated as a second lieutenant and having delayed my active duty service while in graduate school (by now I was a first lieutenant), I had to take a leave of absence after six months at Atlantic to serve my duty tour, which happened to be in the Army Signal Corp's Strategic Communications Command's Computer Systems Directorate, with a good job of suggesting appropriate hardware (like the up-to-date Burroughs B5000) for Army labs and arsenals (getting this kind of technical assignment was in part due to Professor Schultz, who knew just about everybody in industry and government). The Computer Systems Directorate happened to be located in Washington, D.C., which led to my next chance event.

While in Washington, an undergraduate classmate found out that I was stationed there, called me up and asked if I would like to teach a course as a part-time instructor in finite math, linear algebra and statistics (this was all in a single course) at George Washington University (GW). That sounded like fun, and thus I got my first taste of academia.

After finishing my active duty tour, I returned to Philadelphia and to Atlantic, this time working on some inventory problems, as well as site selection problems to determine where to put up new gasoline stations (incidentally, gasoline in the Philadelphia area at that time was 28.9 cents per gallon). As I had enjoyed my teaching experience in Washington, I taught statistics and operations research part time at Drexel University in the graduate engineering administration program. After another year, I realized that I wanted to make a career in academia, and since we had enjoyed living in the D.C. area so much while on active duty, I got in touch with my contacts at GW and was offered a full-time position as an assistant professor in the engineering administration (EA) Department in the School of Engineering and Applied Science (SEAS), where I spent the next 30 years of my career in teaching, research, administration and professional activities.

First, the teaching and research. So, I began my academic career at the tender age of 30 at GW. There were only three departments in SEAS at that time: ours, electrical engineering and computer science (EECS), and mechanical and civil engineering (MCE). Actually, when I first arrived, there were no official departments, just faculty groups, which a few years later became formal departments. EA had only a graduate program — the master's of engineering administration (MEA).

In the EA department of five faculty members, there was one other O.R. type. We proposed an additional master's program in operations research, and a short time later the department changed its name to engineering administration and operations research (EAOR). Since the full-time teaching load was three courses per semester — GW was not a research-oriented university — and I was just starting out and having to prepare for new courses, there was not much time for research.

As luck (chance!) would have it, the university also had an independent (of departments and schools) research program, funded by the Office of Naval Research (ONR), called the Logistics Research Program (LRP), headed by a very able (both in math theory and research management) mathematician named Bill Marlow. About three years after I arrived, this program was merged into SEAS, and Marlow joined the EAOR faculty. This merger resulted in our eventually building up one of the best O.R. departments in the country. Bill Marlow became chair of EAOR.

LRP had many excellent non-faculty computer programmers and analysts and had the largest computer on campus (I think at the time it was an IBM System 360 in the late 1960s and then the IBM 370 in the 1970s — with its "up-to-date" punch card input).

Faculty members from other departments (e.g., Economics, Statistics and Mathematics) worked with LRP on various projects, wherein with sponsored support from LRP, they were able to reduce teaching loads. The same happened for me. I was able to get one-third time support during the academic year, which lowered my teaching load to two courses per semester, as well as two-month's summer support.

A few years after joining GW, I consulted (we were allowed the usual one day per week) for a think tank called the Research Analysis Corporation (RAC), which had previously been the Operations Research Office (ORO) and was supported by the Army. There were many outstanding operations researchers and mathematicians working on Army O.R.-type problems. In the early 1970s, RAC had a downturn in Army funding. At the same time, due to some university politics (we fit right in with the D.C. culture), our dean was able to create some new positions for our department. This allowed us to hire some of the top O.R. professionals in the country: Tony Fiacco and Garth McCormick (of the sequential unconstrained minimization technique fame [SUMT]); Jim Falk, also an outstanding math programmer; Carl Harris (who was to become a close friend and co-author of many papers and a queueing text); and ultimately Rich Soland, a person of many talents including multi-criteria decision theory. All these people had been at RAC during my consulting time there, so when positions opened up, we took advantage of the opportunity. We were also fortunate to attract Nozer Singpurwalla, an outstanding young researcher in reliability theory and statistics, and some years later another young teacher/researcher in stochastic process, Doug Miller.

Thus, by the mid-1970s, we had a critical mass to build up a high quality O.R. department, and the EAOR department split into two: an O.R. department and an EA department. The first chair of the new O.R. department was Bill Marlow, and we set up both M.S. and Ph. D. programs in O.R.

With our expanded faculty, especially the expertise we built in math programming, I was now free to concentrate on my first love — chance! I worked with Carl Harris in developing stochastic processes courses, particularly inventory control and queueing theory, as well as the second semester of a two-semester introduction to O.R. course, which all students had to take — the first being deterministic (math programming) and the second stochastic (queueing, inventory and reliability). We also developed a course in discrete-event simulation and advanced courses in these topics.

Our master's and Ph.D. programs flourished in the 1980s, as did our sponsored research.

The professional societies. One of the things I did shortly after going into industry was to join the Operations Research Society of America (ORSA) and The Institute of Management Sciences (TIMS), the forerunners of INFORMS. These were separate organizations in those days (the 1960s), so one could go to four national meetings a year. I did go to as many as I could, and found these to be wonderful experiences. The top people in the young O.R. field were there, names such as Dantzig, Charnes and Little. Wow! I actually met John D.C. Little of Little's law (remember I was enamored with queueing theory). It was as if a young mathematician went to a math society meeting and met Liebnitz or Pascal, for example. Even though TIMS was generally considered to be management school and industrial practitioner-oriented, and ORSA more engineering, math and military practitioner-oriented, by the 1970s, the two societies were doing many joint endeavors (e.g., national meeting and some publications).

When first arriving at GW, I became involved with the Washington Operations Research Council (WORC, now WINFORMS) and through my work there was asked to be general chair of the 1980 ORSA/TIMS fall national meeting. There was still considerable competition between the two organizations, e.g., the fall joint meetings were ORSA/TIMS and the spring joint meetings were TIMS/ORSA. As luck would have it, I picked great meetings helpers (program chair, arrangements chair, etc.) and it was the most successful meeting to that date. I was elected to ORSA council, joined the ORSA and the joint ORSA/TIMS meetings committees and eventually became chair of these and was elected president of ORSA in 1988. By that time, most activities were joint (and there were not just duplicate but triplicate committees, e.g., an ORSA meetings committee, a TIMS meetings committee and a joint ORSA/TIMS meetings committee, as well as three councils: ORSA, TIMS and ORSA/TIMS).

There had been talk from time to time after the societies started joint activities to merge into one professional society. But loyalties to each group kept this only in the discussion stage for many years. However, as chance would have it, during my tenure as ORSA president, Bill King was elected president of TIMS. Both Bill and I were of a merger mind, and we set up the Little (yes, John D.C. Little of queueing [and other] fame) committee to look into merging the two societies. There were many meetings, and through John's ability and diligence, merger finally came about seven years later in 1995. Many meetings in those seven years were in Cambridge, Mass., at John's MIT office, topped off by a marvelous lobster dinner at his house, cooked humanely by John who cooked the lobsters in beer so they died happy drunks.

My experiences with ORSA, TIMS and INFORMS have been among the most rewarding activities of my career. We embarked on an effort of bringing O.R., math and science to the high school and middle school levels, putting out several videos, developing programs for high-school and middle-school teachers at the national meetings (first started by Ben Lev at his Philadelphia meeting) — all under the able leadership of Col. Frank Trippi, who was then the chair of the INFORMS Public Awareness Committee. Having the honor of putting on a national meeting, being local committee chair of an IFORS meeting, serving with great folks for many years on the ORSA and INFORMS councils, and, of course, the eventual merger into INFORMS are memories I will cherish for the rest of my life.

After having served as department chair for 13 years and associate vice president for research and graduate studies for two years, in the summer of 1995, I took advantage of a very attractive early retirement option from GW and joined, as a research professor, the relatively new School of Information Technology and Engineering at George Mason University, where several of my former colleagues and Ph.D. students were now faculty members. My longtime colleague, research partner and co-author of many papers and a book, Carl Harris, was chair of the O.R. department, and Dean Andy Sage offered me the research appointment in Carl's department. These were among the happiest years of my career, working with former and new colleagues, and having had my last two bosses (department chairs) being former Ph.D. grads of our GW department. We worked on queueing applications in telecommunications and air traffic control (you will notice that there is still lots of room for good research in both of those areas). And, of course, our great fun during the Mason basketball Final Four run in 2006. But happiness is often colored with some sadness, and the sudden death of Carl Harris in 2000, while exercising at his local gym, was a devastating blow to us all at Mason. But in all, it was a marvelous 50 years.

So in closing I say, with all due respect to the optimization crowd, hooray for chance!



Former ORSA President and INFORMS Council Member Donald Gross (dgross1@gmu.edu) and his wife of 48 years, Alice, are enjoying retirement together in Charlottesville, Va.

Table of Contents

OR/MS Today Home Page


OR/MS Today copyright © 2008 by the Institute for Operations Research and the Management Sciences. All rights reserved.


Lionheart Publishing, Inc.
506 Roswell Rd., Suite 220, Marietta, GA 30060 USA
Phone: 770-431-0867 | Fax: 770-432-6969
E-mail: lpi@lionhrtpub.com
URL: http://www.lionhrtpub.com


Web Site © Copyright 2008 by Lionheart Publishing, Inc. All rights reserved.