K

M&S-related Education

Donald Gaver, Naval Postgraduate School

BACKGROUND

The panel recommends an increased effort to educate future officers (and civilians) for work involving modeling and simulation (M&S) (discussed in the text of the report). Some of the work will involve developing M&S. Some will involve applying it. The applications will be in acquisition, training, and operations—each of these broadly construed. In many cases, the applications will be “analytical” in some sense —e.g., investigating the potential value of a new weapon system or tradeoffs among platforms, constructing a training activity or exercise that will expose participants to the desired range of situations and stresses, or assessing alternative courses of action. In other cases, the work will be more developmental or technological (e.g., managing a program that includes a model-building component or managing the assembly of a distributed interactive simulation specifically tailored to an exercise). The purpose of this appendix is to discuss the subjects to which students might be exposed to prepare them for such activities. 1

DISTINGUISHING AMONG CLASSES OF EXPERTISE

If we contemplate the range of military or civilian professionals who will be working extensively with M&S, it quickly becomes evident that there are some distinct specializations. One useful breakdown from an M&S-centered perspective (one partly motivated by the discussion in Chapter 6 about layered architecture for M&S) is as follows:

1  

This appendix has benefited from inputs by Bernard Zeigler and Paul Davis.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 239
Technology for the United States Navy and Marine Corps, 2000-2035: Becoming a 21st-Century Force K M&S-related Education Donald Gaver, Naval Postgraduate School BACKGROUND The panel recommends an increased effort to educate future officers (and civilians) for work involving modeling and simulation (M&S) (discussed in the text of the report). Some of the work will involve developing M&S. Some will involve applying it. The applications will be in acquisition, training, and operations—each of these broadly construed. In many cases, the applications will be “analytical” in some sense —e.g., investigating the potential value of a new weapon system or tradeoffs among platforms, constructing a training activity or exercise that will expose participants to the desired range of situations and stresses, or assessing alternative courses of action. In other cases, the work will be more developmental or technological (e.g., managing a program that includes a model-building component or managing the assembly of a distributed interactive simulation specifically tailored to an exercise). The purpose of this appendix is to discuss the subjects to which students might be exposed to prepare them for such activities. 1 DISTINGUISHING AMONG CLASSES OF EXPERTISE If we contemplate the range of military or civilian professionals who will be working extensively with M&S, it quickly becomes evident that there are some distinct specializations. One useful breakdown from an M&S-centered perspective (one partly motivated by the discussion in Chapter 6 about layered architecture for M&S) is as follows: 1   This appendix has benefited from inputs by Bernard Zeigler and Paul Davis.

OCR for page 239
Technology for the United States Navy and Marine Corps, 2000-2035: Becoming a 21st-Century Force Applications-oriented professionals (modelers, analysts . . .) working in particular application domains such as acquisition, training, or operations, who can effectively pull together M&S assets as needed for their problems. In some cases, they will build or substantially alter models themselves; in other cases they will use preexisting models. They may or may not (and usually will not) consider themselves M&S specialists, often preferring to be identified as analysts, but, at least in the future, will have considerable M&S expertise. Program managers in domains that require overseeing applications-oriented M&S development or model-supported analysis. Hardware-centered professionals trained in the technologies of computers, networks, and related communications. Software-centered professionals trained in developing the software that utilizes the hardware to support the applications. Here we have in mind professional-quality software, not the computer programs typically generated by analysts or subject-focused modelers. Such software is intended for broad use, not just that within the originating group. M&S facilitation specialists trained to draw on technology and databases for both development and application of M&S in relatively complex contexts such as distributed interactive simulation, or developments exploiting model and tool repositories. These specialists would also be experts in assisting the collaboration with other professionals using groupware technologies of all types. M&S scientists, responsible for researching the architecture of both local and distributed M&S infrastructures, continually assessing their capabilities relative to future needs. These might be concerned about n-th generation “high-level architectures,” complex computer-security issues, and tool development. The focus in this appendix is on the applications-oriented professionals and program managers, not because they are more important than the others, but because it is here that the Department of the Navy probably wants to focus its special M&S-related education that is keyed to young officers. In contrast, the Navy Department will probably go to civilian employees or contractors for specialized skills in hardware, software, and so on. There will probably be an adequate supply of people with such skills, people who will have attended colleges and universities throughout the nation. EDUCATION FOR FUTURE M&S USERS A Perspective to Guide “Requirements” Modeling and simulation (M&S) is one tool for use in making effective military decisions. A helpful planning or decision-assisting model is one that captures the essential elements of a situation or problem domain and that can be manipulated to provide synthetic experience efficiently. That experience is then

OCR for page 239
Technology for the United States Navy and Marine Corps, 2000-2035: Becoming a 21st-Century Force used as one input to guide choices of assets, tactics, or policy. An effective training model (or simulation) also provides synthetic experience, but now aimed to allow an operator or team to achieve and maintain particular skills. Parenthetically, decision-assisting models can provide invaluable training for decision makers, especially if either historical or hypothetical situations are presented that illustrate the realistic effects of uncertainty in its many aspects on the decision-making environment, and consequently on subsequent decision-affected outcomes. If the above is an acceptable if abbreviated overview of the M&S enterprise, then one can ask for the background, sensitivities, and expertise desirable in a well-prepared professional user of M &S. The first observation is that today all such properties are unlikely to be embodied in one individual. What follows is a suggested order of priority for the types of talent and experience needed when an M&S enterprise is to be pursued. This list helps to define the educational needs. Designer-Architect-Problem Formulator No very substantial project involving M&S should be initiated without articulating one or more specific issues or questions to be examined. These questions should relate to the purposes of the organization guided by the decision maker to be advised, and should be as focused as possible. It requires art and experience to identify such questions; skill comes with practice. This arena is the purview of the essential designer-architect-problem formulator, whose proposals and direction set the stage for subsequent more technical modeling steps. There is scattered literature useful to educate such specialist-generalists; some classics like G. Polya's How to Solve It are useful to read, but “how to formulate it” is more to the point, and some intensive searching for and creation of useful teaching material are in order. There are a number of books and courses somewhat relevant to such matters, sometimes in an operations-research or policy-analysis curriculum, sometimes elsewhere. Experience suggests that the case study approach is particularly valuable, because one learns how to formulate and conduct studies more by doing than by merely hearing principles. Accessing historical examples is a natural way to proceed; mining and refining corporate memory in particular areas and organizations can be undertaken to record lasting “lessons learned.” Students can also be tasked to conduct “quick-response studies,” which can be effective in instilling recognition that much can be done quickly with a mix of brainstorming, simple or relatively simple models, and clear problem-focused thinking. On a formal-training level an initial educational background of natural science such as physics, chemistry, and electrical engineering, but also applied mathematics and statistics, has often been useful, particularly if the individual “likes problems” and enjoys the uncovering and exploitation of obscure structure and mechanism. Courses in mathematical and applied probability modeling, if designed

OCR for page 239
Technology for the United States Navy and Marine Corps, 2000-2035: Becoming a 21st-Century Force around problems (i.e., rather than being designed to illustrate sophisticated mathematical techniques) can be effective. Project work and informed mentoring are key to training new practitioners. Delivery of formal education or training in this essential M&S function has not been well addressed in very many places. Conceptual Modeling Once the desired questions are framed, choice of formal representation(s) of the problem elements must be made. It is often good practice to maintain several alternatives; for example, an initial low-resolution but fast and agile model could be explored, subsequently selectively enhanced by a more detailed higher-resolution followup, or a deterministic approach could be followed by a stochastic version. It is also valuable for students to be exposed to different perspectives of “the same domain,” perspectives such as entity-level simulation on the one hand and operations-research analytical models to guide resource allocation on the other. A goal here should be to teach students to recognize and appreciate the values of different perspectives and representations, rather than associating themselves emotionally and nearly exclusively with one or the other. The student is traditionally made aware of a number of model-type tools, typically quantitative-mathematical in nature but more recently also visual and animated, particularly in the training arena. Parenthetically, there is complementary overlap: statistical use is being made of “data animation and visualization” for dramatically conveying messages buried in complex data structures. Such can also be done to expedite model exploration. Classical Methods Here are some traditional model types and modeling tools that M&S professionals should know and appreciate to varying degree (no single individual is likely to be deeply conversant with all): Mathematical programming and optimization; other search methods such as evolutionary programming or genetic algorithms; Probability models and stochastic processes; search theory; reliability models; queuing theory; Statistics: data acquisition, and data analysis; Spreadsheet languages; simulation languages, others; Monte Carlo methods; Decision and control theory and analysis; Artificial intelligence (AI); rule-based systems; knowledge-based simulation; Game theory, game-theoretic optimization in simulations, and adversarial knowledge-based models in simulations (especially under uncertainty);

OCR for page 239
Technology for the United States Navy and Marine Corps, 2000-2035: Becoming a 21st-Century Force Dynamical systems; chaos and complexity ideas; cellular automata; and Human factors and performance, and human-computer interfaces. Higher-level Issues of Design An important subject that is not often taught well is model design, especially for complex systems. One recently developed subject that is quite relevant here is object-oriented modeling (as distinct from programming), including modeling of complex systems. Students can profit from a study of systems dynamics, a subject usually associated with MIT's Jay Forrester. A classic book that addresses the nature and modeling of complex systems is Herbert Simon's Sciences of the Artificial (Simon, 1996). New subjects of considerable importance involve complex adaptive systems and agent-based modeling. There are some interesting popular and semipopular materials available, but no single best reference of which we are aware suitable for graduate education. Appendix B discusses many of the items of interest here. Appendix E emphasizes the importance of multi-resolution modeling and the desirability of having model families, but there is very little in current curricula — and not much in the current literature—to prepare people for such work. It will often be true that a real problem can be completely addressed or “solved” by employing some relatively simple classical model. This possibility should not be overlooked. Model Choice and Adaptation This is the stage at which computer-intensive tools are invoked and computer science ideas find a place. Some important topics in this domain include the following: Object-oriented programming and, more generally, software engineering; Computer architecture and operating principles; Computer and communication networks; security; M&S tools and practices to aid comprehensibility, traceability, and “explanation”; M&S designs to encourage and facilitate “exploratory analysis” amidst great uncertainty; Virtual-world and simulation systems; Distributed (operating) systems; and Virtual reality, and distributed interactive systems (DIS). The M&S practitioner may well be required to address problems with versions of existing, even Service-specific models, such as the Navy's ITEM (or NSS), Army VIC or EAGLE, or Air Force THUNDER, or higher-level joint

OCR for page 239
Technology for the United States Navy and Marine Corps, 2000-2035: Becoming a 21st-Century Force systems such as TACWAR and JICM, and JWARS and JSIMS when they become available. He must be able to adapt these to particular situations and questions, being critical concerning results obtained. This step also includes basic parameter specification. Model-output Analysis: Analytical Advice to Decision Makers Model-output analysis is an extension of the above that includes the planning of model runs so as to economically obtain the necessary overall picture of response possibilities. Some of the above should be analyst-induced or instigated, while some can be in dialogue style with decision makers. In summary, the professional user of M&S is desirably, but not currently realistically, responsible for a broad spectrum of knowledge and skills. This requires intensive, specific, and well-designed educational input with deliberate breadth and focus on the true usefulness of various viewpoints and technical tools. The field is bound to grow, and competition for appropriate military analysts so trained will grow also, but the opportunities should attract high-quality students and prospective practitioners. VENUES FOR EDUCATION Most of our discussion here relates primarily to postgraduate education in universities (e.g., in master's or Ph.D. programs), but an increasingly important part of educational strategy for organizations such as the Department of the Navy is the part that makes available specialized courses on an as-needed basis—e.g., before an officer takes on an assignment overseeing M&S development, exercise design, or weapon system analysis. Many possibilities exist here, ranging from short courses to self-learning packages. They are not complete substitutes for traditional degree studies, but they can be powerful supplements and, in some cases, partial substitutes. Continuing education is becoming a business-as-usual aspect of life for many knowledge workers, whether in or out of uniform.