This article presents a case for a growth of the discipline by encouraging systematic collaboration via the formation of consortia comprising of members from academe and industry. These consortia will provide a forum through which members interested in specific sectors of the economy (e.g. airlines, energy, etc.) may work to develop realistic scenarios and raise O.R. questions that will enhance industrial competitiveness under these scenarios. In turn, such collaboration will enhance the applicability of O.R. research.

In the basic sciences (physics, chemistry, biology, etc.), and even some social sciences, researchers not only propound new theories, but also verify how well these theories predict observations in the real world. In the long run, theories with better predictive performance survive, and others disappear. O.R., which traces its origins to physics, mathematics and statistics, used to be such an academic discipline. The finest O.R. professors had an interest in verifying whether their ideas had immediate practical value. The late George Dantzig, Abraham Charnes and others were not simply talented theoreticians; they took their theories to the real world and demonstrated the value of their models and algorithms. While some O.R. professors do continue in this tradition, a significant proportion of research in academe will never be tested with real-world data.

O.R. in Practice


Over the years, the INFORMS Edelman competition has provided the worldwide stage for innovative O.R. practice, and as a testament to the growth of our field, the competition has become stiffer over the years. Thanks to Friedman's best-selling book (Friedman [2005]), the popular press has also awakened to O.R. success stories such as the logistics revolution, the role of O.R. in UPS, etc. The fact that the algorithm underlying Google searches was inspired by Markov Chains should be enough to convince skeptics that O.R. remains a magnificent source of bright ideas. Google was not the first search engine on the market, but O.R. ideas helped them develop a better search engine.

O.R. has played a significant role in advancing both knowledge and the modern way of life. Nevertheless, the burden of industrial innovation has been borne primarily by practitioners in industry. This is not to suggest that there are absolutely no cases of interactions between the real world and O.R. academe. A recent editorial by Rothkopf [2005] provides a comprehensive look at the involvement of academe in papers regarding O.R. practice. It is remarkable that over a seven-year period (ending in 2005), only one university had more than 10 practice publications!

The point is that interaction between O.R. academe and industry (or at least the published record of it) is minimal. Because of such minimal interactions, our curricula, in general, have a far greater emphasis on the conceptual underpinnings of O.R. than its practice. Consequently, O.R. graduates hired into industrial consulting and research positions find themselves at a disadvantage in making the transition from academe into practice. The need for greater participation in real-world innovation by the academic community is clear. This will not only elevate O.R.'s impact on the real world, but also provide corporations a competitive edge.

The Barriers


As outlined by Lougee-Heimer et al in the accompanying article (see page 34), there are a variety of barriers to connecting academic research to practice. I will only touch upon what I believe is the main hurdle.

Many major U.S. corporations (e.g. Boeing, Ford, General Electric, General Motors, Hewlett-Packard, IBM, Intel, Lucent, UPS and others) have long recognized the value of O.R. in enhancing their competitive positions. The economic value of novel O.R. methodology, together with a well-trained cadre of experts who can bridge

academic research and industrial innovation, is enormous. As argued by Friedman [2005], the economic landscape affected by O.R. includes businesses of all sizes: small firms are able to piggyback on "logistics suppliers" to compete on a global scale, and large, global enterprises are being redesigned for greater agility. Several service-oriented corporations have thrived because they have been

able to gain the trust of their clients and have formed fruitful partnerships in which collective success is the only viable outcome.

On the other hand, university research and industrial data remain separated by a gulf of legal barriers. With some exceptions, defining the "intellectual property" agreement (between a corporation and a university) that enables a corporation to engage academic O.R. research talent and provide access to real data is a roadblock to collaboration. As a result, academic research in O.R. has retreated further into the ivory tower than envisioned in the early days of the discipline. Consequently, significant opportunities for challenging research are overlooked by academe, and popular academic models (with few industrial connections) continue to engage O.R. research talent. On the industry side of the equation, the lack of research connectivity leads to longer development cycles, higher development costs and slower responses to pressing problems. Instead of creating a "win-win" proposition through collaborative research, the widening separation between academe and industry may stifle the growth of the enormously powerful O.R. discipline.

Bridging the Gap


COIN-OR's Real Problem Exchange (RPX) (see page 34) is a first step in bridging the gap. Such an exchange will allow the industry to pose challenging problems that, in turn, will motivate academe to develop new models, algorithms and software. While the exact nature and role of such an exchange is not clear at this time, it is clear that RPX has the potential to revolutionize the way in which O.R. models and algorithms are validated. At a very basic level, one can think of RPX as simply a platform for instances such as those available in LPLIB, MIPLIB, SIPLIB, etc. Another form of a "problem" is the famous "beer game" for inventory management.

One can easily envision a far more powerful exchange environment, however, that would enrich O.R. research by providing a variety of simulations that bring together machines, humans and markets, i.e. the real world. For instance, a challenging financial engineering "problem" may be posed in the form of a market simulation which not only shows the user (say a Ph.D. student) a trajectory of returns (due to a particular strategy), but also provides the user an opportunity to query a client's perception of performance. Within such an environment, the value of new O.R. research may be judged on its merits within the simulated world.

The greater the fidelity of the simulation, the greater the justification for evaluating the practical merits of the conclusions (from the simulation study). It is clear that setting up such simulations can themselves pose interesting research questions because of the need for multi-scale, modular simulations. Even more apparent is the need to bring O.R. academe and industry together to pose significant research "problems." One consortium that has had a long and fruitful history of such an effort is SEMATECH. However, most other sectors of O.R. applications have not developed such synergy.

My suggestion is to encourage INFORMS sections and societies to facilitate the creation of consortia consisting of academic and industrial members. These consortia will be charged with developing scenarios and posing problems that neither industry nor academe is able to address today. These problems should be such that an entire sector of industry must foresee the need for greater insights provided by new models and algorithms. By describing these scenarios and associated problems via a generic simulation, no specific firm will have to reveal its specific data, thus overcoming one of the major barriers. I am certain that this is an oversimplification of the issue of getting more realistic applications to motivate O.R. research. Nevertheless, creating such consortia can only help the discipline better serve its constituencies in an increasingly competitive world.

Acknowledgements I would like to thank Robin Lougee-Heimer for many discussions on the topic of this article and her leadership in getting RPX started within COIN-OR. I am also grateful to Abhi Deshmukh, Debra Elkins, Bob Fourer, Michael Grant, Dave Heltne, Ben Lowe and Ranganath Nuggehalli for sharing their views at the INFORMS meeting in San Francisco ("Creating a Testbed of Industry Problems") in November 2005. Finally, I would also like to thank Stephen Nash for his comments on an earlier version of this article.

1. Friedman, T.L., 2005, "The World is Flat," Farrar, Straus and Giroux, New York
   
   
2. Lougee-Heimer, R., et al, 2006, "Exchange Keeps It Real," OR/MS Today, Vol. 33, No. 1, pp. 34-37.
   
   
3. Rothkopf, M.H., 2005, "The Sixth Interfaces Ranking of Universities' Contribution to the Practice Literature," Interfaces, Vol. 35, pp. 425-428.

Table of Contents

OR/MS Today Home Page


OR/MS Today copyright © 2006 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 2006 by Lionheart Publishing, Inc. All rights reserved.