In putting together a representative series of papers about Military Operations Research, we quickly realized that in many ways we are putting together a survey of both the past and future of our profession. So many of the tools and techniques that are ubiquitous in the general practice of operations research (O.R.) come from Military Applications. The first use of the term “Operational Research” appears from the British effort during the Second World War¹. The seminal text by Morse and Kimball “Methods of Operations Research” (US Navy Operations Evaluation Group (OEG) Report #54) focuses, as its subtitle implies – on military applications. Other cornerstones of our profession, such as “linear programming,” as developed by Dantzig, refers to the creation of a budgetary Program for Defense (not lines of computer code²), and the term Monte Carlo simulation traces back to the Manhattan Project.³

As we set to work, we found ourselves asking “What is the definition of military operations research?” While saying “The application of operations research to military problems” is true, it is unsatisfying as it simply restates the words. We immodestly rephrase Morse and Kimball’s classic definition of O.R. to state that military operations research is: “The application of quantitative analytic techniques to inform military decision making.” To us, the distinguishing attributes of Military Operations Research are the following:

1. It is, fundamentally, applied in nature. There has been some excellent theoretical work as part of Military O.R., but the focus of Military O.R. is always a specific problem and the end result is an answer that can be implemented by a decision maker in a relevant time period.
2. At its best, it is problem focused versus tool focused. Individual practitioners naturally have focus areas – your editors’ are stochastics and combinatorial optimization. But the O.R. community follows the problems where they lead, even if that is areas outside the practitioners’ general (or preferred) fields.

In synthesis, these two attributes make for a diverse array of applications and methods, as evidenced by the collection of papers to follow.

In the early days of operations research, military problems driven by the Second World War, Cold War, and coincident competitions such as the Arms Race and the Space Race preceded and catalyzed the development of civil analogues. For example, the max-flow min-cut theorem was developed in the context of preventing Soviet armies from reaching the Fulda Gap via road and rail, but has since been applied to countless civilian network flow optimization problems; you probably experienced its impact if you have flown on a commercial airplane or purchased goods from a major retail chain. The algorithms and techniques used to find submarines (or insurgents) are also used to find missing vessels and aircraft at Sea⁴; the large-scale dynamic methods and computing environments used to simulate nuclear explosions are also used to predict the weather. In this sense, the history of Military Operations Research truly has examples of ‘Swords beaten into Plowshares.’

Over time and as the problems have changed, this lead-follow relationship – as measured by the pace of innovation – has become more balanced. The techniques originally developed in the context of military conflict have found additional applications in our everyday lives, but in more recent times the military O.R. community has – arguably – struggled to keep pace with tech innovation in the civilian world, particularly with Silicon Valley companies. Whereas as in the late 1940s and 1950s, the most cutting-edge O.R. techniques were developed by the (US) Department of Defense (and others) and classified out of the public domain, today the Defense Analysts worldwide are making a concerted effort to keep pace with and leverage the latest commercial technologies through organizations like the Defense Innovation Unit Experimental (DIUx).

The Volume

Our purpose with this Editor’s Cut is to provide a primer to those new to the field of military O.R., and – if we may be so bold – to establish a touchstone of what we believe to be truly great, even classic, examples of military O.R. done right.

This volume focuses on published papers, and therefore excludes books from consideration; INFORMS has compiled an excellent list of books⁵; a quick perusal will show how many of these books have military applications in their DNA.

In selecting papers for this volume, we tried to balance the strong sense of heritage against cutting edge techniques. Also, as an INFORMS publication, this volume naturally features INFORMS papers; there are other outlets where military O.R. is featured, specifically those from the Military Operations Research Society (MORS), of which your editors have both served as directors. Finally, there are independent monographs, produced by independent institutions⁶; we present a good introduction but by no means the complete story.

You will see many disciplines represented here – statistics, optimization, cost estimation, manpower planning, and deterministic methods. In selecting papers, we considered three criteria:

1. Impact as measured by translatability to applied work; the ability to deliver defensible answers to operational questions.
2. Recent papers were preferred over older papers.
3. Papers that fold in advances from other fields – such as computing – were preferred.

In addition to technical papers, we have included a few retrospectives and philosophical tidbits that we believe to be useful to the community.

The Future

As we look forward, we nominate some challenging problem spaces that are ripe for future work and nascent technologies that may help solve them:

1. A better understanding of the role of autonomy in all aspects of military operations, from the dynamic to the mundane. Specifically, a deeper exploration of what machine learning and artificial intelligence can – and cannot – contribute to these problem sets.
2. A better understanding of the role of cyber and information operations in modern warfare, particularly when balanced against competing requirements for traditional military systems and forces within no-growth or shrinking defense budgets.
3. A renewed understanding of both tactical and strategic deterrence, building on classic ideas, but revamped for the information age, as well as a deeper, empirically-based understanding of the role of nontraditional and nonstate actors.
4. The development of new technologies with high potential to disrupt quantitative analysis and computation, especially quantum and neuromorphic computing.

Concluding Remarks

We hope that newcomers find this volume to be a broad (if necessarily brief) introduction to the field, and that even seasoned professionals might find some new gems from outside their practice. We also hope that this volume will serve as inspiration for O.R. professionals from outside the defense space.

Notes

1. https://www.informs.org/Explore/History-of-O.R.-Excellence/Bibliographies/The-Origins-of-OR
2. https://cs.nyu.edu/overton/g22_lp/encyc/article_web.html
3. Metropolis, N., & Ulam, S. (1949). “The Monte Carlo method.” Journal of the American Statistical Association, 44(247), 335-341.
4. https://www.informs.org/ORMS-Today/Public-Articles/February-Volume-42-Number-1/Aviation-Safety
5. https://www.informs.org/Publications/Topics-in-O.R/Book-List
6. A prime example of which is “Thinking About America’s Defense,” https://www.rand.org/pubs/occasional_papers/OP223.html.