This report lays out important improvements to practice, research areas, and educational requirements needed to address emerging MS&A and military challenges. Developing an in-depth understanding of the challenges and the solutions requires updating and strengthening the base of military science, in addition to the more straightforward refining of technology and tools.
DEPARTMENT OF DEFENSE’S NEED TO LEAD IN MS&A AND MILITARY SCIENCE
In Chapters 2 and 3, the committee recommends substantial effort and investment to build the base of M&S science to deal effectively with networking; embedded systems; the implications of interconnected complex, dynamic, and adaptive systems; and other emerging issues. The committee identifies numerous topics and approaches that appear to it to be important for this enterprise. It will be essential, the committee believes, for DoD to approach this research, and subsequent applications, in a scientifically and analytically sound way. Doing so will require something that does not yet exist—one or more organizations in DoD (including the uniformed services) that have a specific charter to encourage, nurture, and exploit the results of the research and changes that the committee recommends. These organizations need to be staffed with individuals having a high degree of technical competence, an understanding of military issues, and a passion for organizing and codifying knowledge (or helping to stimulate such actions).
DoD will need to nurture and sustain such fundamental advances, and it should identify which organization(s) should have this responsibility as part of their charter. As with much research, such investments should be seen as “common goods” justifying centralized support, whether by a service, a joint organization, or the Office of the Secretary of Defense. Although the Defense Modeling and Simulation Office (DMSO) has produced and supported a great deal of infrastructure development—for example, that of the High-Level Architecture (HLA)—it has not regarded investment in the science of MS&A as part of its charter. Either this should change or some other organization should be tasked with such investment and nurturing.
Research in MS&A is needed to improve the modeling of real combat, and the organization(s) charged with leading this effort should go one step further and lead the development of basic military science. By this is meant the study of the technical, psychological, practical, and other phenomena that constitute warfare and armed conflict.
Military operations today comprise a large number of interconnected and counterposing systems and variables that, in addition to characterizing kinetic and other direct modes of confrontation, include special operations, combat service support, logistics, medical support, intelligence, counterintelligence, psychological operations, etc. Noncombatant entities typically outnumber combat entities, and more now needs to be learned about noncombatant activity than traditional direct force-on-force activities.
Such a new science of military operations would by necessity incorporate the reality of the civilian presence in many forms, from informants during combat operations to insurgents during reconstruction phases. This makes it difficult to represent phenomenology and, most contentiously, to report outcomes or sequelae of engagement with even the most sophisticated of current MS&A capabilities. Moreover, it is reasonable to argue that activity taking place in the so-called criterion environment itself—i.e., the battlespace—is so complicated that it is an inherently unreliable target for validating combat models. The systems being modeled may be highly nonstationary as well as stochastic. If that is so, then the usefulness of conventional stochastic models could be severely limited. This so-called “validity problem” represents a formidable—even (some would argue) overwhelming—problem for today’s military science base.
Such deterrents to representing the most straightforward, contemporary military problems mean that, arguably, the military science base is exploited today well beyond what
can be justified. Moreover, as the battlespace continues to increase in complexity, and as outcomes rely less on mere attritional or maneuver success and more on multivariable interactions, new computational approaches will be needed to support the more complex MS&A required. This difficulty is exacerbated by MS&A’s inability to represent network-centric phenomena.
Combat is viewed by many today as systems of interconnected systems or, more formally, as complex adaptive systems (CAS). These CAS generate properties that are not unlike those that have been examined over the past decade or so by scientists who are concerned with complex phenomena in economics, meteorology, and ecosystems and who claim to have the computational capability to do so adequately. Salient among these complex, adaptive properties are self-organization and emergence.
On the battlefield, “exploitable complexity” is the military’s attempt to produce self-organizing and emergent effects through indirect (often very indirect) means. In fact, the generation of effects-based operations (EBO) is viewed in military science as having the potential to be important in the future, for reasons that are beyond the scope of this report. Nevertheless, in the future MS&A must be able to facilitate the discovery, derivation, and analysis of tactics, operations, and strategies that produce “intended emergence,” which is precisely what EBO attempts to do. There are arguably few problems that are more important, or more difficult, to represent than those associated with EBO. To provide the necessary analytic support for EBO, it is essential to significantly accelerate the rate of progress in the entire science base underlying military MS&A, including both MS&A science and military science.
A program of research for military ends, including both MS&A science and basic military science, is unlikely to happen without DoD sponsorship and support. The area is not particularly attractive to outsiders and, even if it were, they would find it difficult to get access to the needed information. Defense MS&A is crucially important, requiring DoD support in order to avoid losing many of the benefits from DoD’s efforts in the more generic aspects of modeling and simulation.
Recommendation 14: DoD should identify (or create) and charge an organization with responsibility for developing and supporting a program of research and development directed at improving and updating the base of military science for combat and noncombat modeling. That same organization would be responsible for effecting the recommendations on education that are called for in Chapter 5.
To be sure, there are many shortcomings even in the scientific base for traditional MS&A, as well as continuing debates about such things as the appropriate form of attrition equations, movement-rate equations, and so on. The committee does not mean to discourage continuing research in these areas but wishes to encourage the new MS&A challenges where no satisfactory base currently exists.
STEPS FOR ADVANCING MS&A IN ENGINEERING
As noted in Chapter 1, the NSF recently published a report of two workshops on simulation-based engineering science that examined the role of MS&A for U.S. science and engineering, including medicine, materials, and other scientific disciplines. While that report touched on defense applications, it did not bring out the three central themes of the current report—network-centricity, complex adaptive systems, and embedded systems.
The NSF report did, however, identify many of the same issues facing engineering MS&A as does this report, and there are some similarities in the recommendations as well. In particular, the NSF workshops and this committee have some similar conclusions:
NSF. “Formidable obstacles remain in linking highly disparate length and time scales….” The committee expressed DoD’s necessity for overcoming these obstacles in Chapter 3 in the subsection “Multiresolution Modeling and Families of Models and Games” and proposed solutions in the section “Composability.”
NSF. “Verification, validation, and uncertainty quantification are challenging and necessary research areas that must be actively pursued.” The committee comes to the same conclusion for DoD, with a different emphasis, in “Expanded Concepts of Validation” in Chapter 3 and “Addressing Uncertainties” in Chapter 4.
NSF. “Research is needed to effectively use and integrate data-intensive computing systems, ubiquitous sensors and high-resolution detectors, imaging devices, and other data-gathering storage and distribution devices, and to develop methodologies and theoretical frameworks for their integration into simulation systems.” The subsection “Improved Data Collection for MS&A” in Chapter 3 of this report examines this issue in detail, and the section “Building the Scientific Base for Embedded MS&A,” also in Chapter 3, discusses the integration of real-time data into embedded systems.
NSF. “Computer visualization will be integral to our ability to interpret and utilize large data sets….” This agrees with the committee’s finding, expressed in “Visualization of High-Dimensional Data” in Chapter 3.
NSF. “Meaningful advances in simulation based engineering science will require dramatic changes in science and engineering education.” The committee discusses this topic at length in Chapter 5, with a focus on changes needed in the defense environment.
The similarities between the NSF’s assessment of MS&A
and the committee’s own assessment reinforces the committee’s Recommendations 4, 8, and 9—that DoD undertake joint research programs with other government agencies where appropriate, citing network science and embedded systems as two areas where such joint research would be possible.
Despite the similarities, neither of the two reports subsumes the other, but they mutually support the conclusion that MS&A is a vital tool for achieving military superiority and economic competitiveness. Where the substance of the two reports does intersect, there is agreement on the details of how to go forward.
INTERNATIONAL ASPECTS OF MS&A
The committee did not address coalition MS&A in any depth; however, a few observations can be made. Most of the standard MS&A used for training and analysis by both the United States and its coalition partners focuses on the kinetic aspects of warfare. However, other nations are now seriously engaged in pursuing the issues highlighted in this report:
Complexity. The Technical Cooperation Program (TTCP), which links work of Australia, Canada, New Zealand, the United Kingdom, and the United States, has hosted symposia on agent-based modeling and related issues. Much of the innovative research highlighted at these symposia has been done in nations other than the United States. The work of Project Albert, initiated by the U.S. Marine Corps, has also had substantial international participation.1
Command, control, and networking. Important work on the transformational aspects of command and control has been pursued in the NATO context, where, as would be expected, much of the emphasis is on interoperability and on identifying core principles and methods that can be the basis of coalition operations. A recent good example of this is SAS 050, Exploring New Command and Control Concepts and Capabilities, Final Report, prepared for NATO, January 2006, available on the Web site of DoD’s Command and Control Research Program, http://www.dodccrp. org/SAS/SAS-050%20Final%20Report.pdf.
Joint Forces Command. Under its manager for the training of joint and coalition forces and interoperability, the Joint Forces Command has an aggressive program of engaging partners in model-driven exercises. In many cases, U.S. models are used exclusively and U.S. systems extended to represent foreign systems. However, some countries, notably NATO members with a long history of interoperability with the United States, have models of their own. Like the U.S. system, these tend to focus on the kinetic aspects of warfare.
Allied Command Transformation (ACT). This is a subcommand of NATO; more information on it is available at http://www.act.nato.int/. Representatives of the committee visited its Future Capabilities, Research, and Technology (FCRT) subdivision in Norfolk, Virginia. They learned about the M&S activity in progress at the FCRT’s Modeling and Simulation Coordination Section. That section supports ACT in developing and refining concepts, defining capability requirements, and devising experimental and testing programs. Its basic approach is to use commercial off-the-shelf software for concept development and exploration, which is less expensive (in both money and time) than using large, complex existing packages or building new custom software. Some of the commercial packages in use at the time of the visit included Extend, AnyLogic, @Risk, and StatFit. The committee thought that this approach, which is standard in industry but which it had not seen used elsewhere in DoD, had the potential for reducing the time and cost of simulation studies in many areas, thereby increasing the effectiveness of simulation as a management tool.
Some of the themes emphasized in this report, in particular Recommendation 1, which called for developing flexible, adaptive, and robust systems, will facilitate coalition MS&A work. This includes MS&A directly related to operations such as mission rehearsal and command and control. In particular, flexibility, adaptability, and robustness are needed for networking, composability, revised approaches to VV&A and, most generally, for extending MS&A to account for DIME/PMESII issues, many of which depend strongly on alliance or coalitional activities. Further attention to the future of MS&A should, if at all possible, devote considerable effort to international issues.
This report has noted that DoD’s MS&A enterprise is widely distributed across many offices and programs with differing missions and priorities. Although there is value in diversity, there is also value in coordination, direction, and guidance. For that reason, an R&D office, as suggested in Recommendation 14, could benefit all of DoD with the most advanced thinking and insights in this complex area. This would be a natural adjunct to the coordinating role played by the Modeling and Simulation Coordination Office (formally DMSO), and it would create an intellectual common ground for the military MS&A community that would help it address the recommendations in Chapters 3 and 5. Recommendation 14 is addressed to high-level DoD planners.
The committee has concentrated on areas that it regards as underdeveloped but vital for the future of DoD MS&A and in which there is promise of disproportionately large advances. This does not, however, eliminate the need for continued improvement in more traditional areas such as force-on-force modeling, logistics modeling, and transportation modeling. Recommendations 1 through 5 and Recommendation 11 highlight the directions that the committee believes will be most important for the new challenges facing the United States: FAR methods and better methods for modeling network-centricity, embedded systems, and the incorporation of uncertainty. Recommendations 6 through 10 recognize five of the research areas discussed in Chapters 3 and 4 as paramount: social behavioral networks, game-based training and simulation, network science, embedded systems, and cognitive decision making. In Recommendations 8 and 9, the committee suggests that research in network science and embedded networks be undertaken jointly with other agencies. If resources for research are constrained, these five areas should have the highest priority. These first 11 recommendations should be the responsibility of those individuals and organizations that manage DoD’s MS&A research and development.
An educated and effective MS&A workforce is essential for both progress and practice, and in Recommendations 12 and 13 the committee gives guidance for their development. These two recommendations should be the responsibility of the human resource managers in the DoD MS&A community. They also give guidance to individual practitioners as they plan their own careers.
The committee believes that the future of MS&A depends upon progress in both infrastructure and research. It has attempted to give guidance in both areas, tempered by its view of future challenges and likely scientific advances.