Simulation Plots Its Own Path

October 2004
By Robert K. Ackerman
E-mail About the Author


A helicopter designed by developers of the online game America’s Army demonstrates the level of simulation fidelity possible using current technologies. Massively multiplayer online games such as this one may be the future of military simulation.

Dispersal rather than consolidation will guide advanced defense modeling and training.

As defense simulation grows more complex and more capable, it is segmenting just as it moves toward greater interoperability. Instead of diverse simulations evolving into a single, all-encompassing synthetic battlespace, the course is toward individual activities or systems simulated by powerful computing technologies. The goal of modeling developers is to treat these new simulations as modules and assemble them into large-scale simulations that are tailored to trainees’ or commanders’ requirements.

This trend toward creating specialized modules allows simulation experts to plan future systems that offer entirely new approaches and applications. Researchers are looking at Web services as they aim to develop advanced online game-playing that would train large numbers of personnel. Simulations may consist of multimedia systems that actually increase sensory input to humans. And, simulations ultimately may be able to perform cultural modeling, providing insight into local inhabitants’ reactions to events such as military operations in their midst.

Information technology improvements are allowing experts to focus on narrower, but more complex, simulation applications. As these new applications emerge, defense officials would integrate them in various configurations to suit military modeling and training needs. The defensewide simulation architecture would comprise modules that either serve individual needs or connect to meet broader simulation requirements.

However, many hurdles remain to be overcome to achieve this future of defense simulation, according to Dr. Susan K. Numrich, deputy director for technology at the Defense Modeling and Simulation Office (DMSO). A physicist by trade, Numrich is on detail to the DMSO from the Naval Research Laboratory.

“I keep hoping that we have built our last monolithic simulation,” Numrich declares. “We decide where we are going to go by what it is that we are doing. We haven’t learned to sit back and say what should be common and what should be in pieces.”

The modern military is increasing its reliance on sensors as the force moves toward information warfare, evolving command and control, and advanced surveillance and reconnaissance. Planning, acquisition, training and testing, and evaluation all are relying on modeling and simulation to an increasing degree. Many simulation systems are developed independently, so the DMSO is working to ensure that these systems are built around standards that ensure interoperability.

At the core of this effort is a focus on aspects that are inherent in all simulations throughout the military, regardless of originating organization or operational focus. The DMSO provides common pieces for broad-based simulations to provide a joint battlespace environment that can be used by all sensors in a common manner, Numrich says.

“We are more purple, but we are going to have to be technically more integrated,” Numrich states. “We are getting more complex technically, so the integration problem becomes larger. We have a whole host of problems facing us when we try to replicate complexity.”

Each service creates its own elements of the battlespace to carry out its simulation mission. The DMSO is striving to bring these various bits and pieces together into a common arena for reasons that go beyond effective joint training. Numrich notes that, for the department to evolve beyond a platform-centric view to acquisition to a capabilities-centric view, it must represent all of the potential capabilities that might be fulfilling a particular role—and, it must accomplish this in the same arena. All of these simulation parts must function together meaningfully, she adds.

Several common pieces are inherent in all military simulations, Numrich warrants. These elements of the joint battlespace environments effort include equipment types, equipment groupings and capabilities, the environment and human performance.

The services’ disparate approach to simulation has created some conflicts among these aspects that the DMSO must rectify, Numrich points out. For example, U.S. Army simulators could have a definition of the ocean that focuses largely on surface water effects. However, the U.S. Navy’s ocean definition would include sub-surface conditions applicable to submarine operations.

Keeping these diverse military simulations on the same page requires standardizing the data. So, to deal with environmental aspects, the DMSO has a naming convention that has come out of a program known as Synthetic Environment Data Representation and Interchange Specification, or SEDRIS. This system has two key functions: representation of environmental data and the interchange of environmental data sets. These functions allow the DMSO to collect data and to name it in an orderly manner that permits simulations to recognize the data semantically and practically.

After the data is named, simulators must be able to find it. A master environmental library allows them to do that. If the data does not exist, an environment scenario generator capability takes the past 40 years of actual weather conditions and generates it in a geospatially correct form. Numrich relates that this data is generated at the resolution, and with the environmental factors, needed by the user.

Numrich notes that the DMSO has executive agents for each environmental discipline. The National Geospatial-Intelligence Agency serves as the terrain agent, the U.S. Air Force deals with atmosphere and space, and the Navy is in charge of the ocean environment.


The Close Combat Tactical Trainer is a distributed interactive simulation that can be linked with other simulations to provide a multifaceted model of combat vehicle operation. Increased specialization in simulation is leading toward more modular modeling in a distributed setting.

Among other executive agents is the Defense Intelligence Agency, which certifies representations of the enemy in the category of human performance. This definition is confounded by new variables such as the presence of civilians and complications arising from people being immersed in different cultures.

The DMSO has “significant simulation programs” in human performance, Numrich allows. These encompass factors such as how humans will react against nonlethal weapons and how crowds react to a particular situation. Many of these conditions are emerging in Iraq, where enemies often are indistinguishable from innocent civilians, and military forces can benefit from training in understanding the dynamics of a crowd.

“In a simulation, we will try to represent those capabilities so that users can work with things such as rules of engagement,” Numrich says. “Young, dedicated men and women are sitting at crossroads judging whether or not someone or some group has lethal intent. We are hoping to put that kind of behavior into a simulation environment that will be both engaging and instructive for them.”

The future promises even more challenges for simulators. For example, the ongoing force transformation that is revolutionizing the U.S. military is spawning successive generations of new types of equipment. In many cases, simulation experts cannot even envision the new hardware and software that they may be simulating in less than three decades. “Some things that are part of a gleam in the back of an engineer’s mind will be part of our military in 20 or 30 years,” Numrich observes.

“We have to be able to represent those pieces, and the only way you can actually look at them is in a synthetic battlespace,” she adds. 

One factor that is helping military simulators is that the government is not the only organization building these systems. The commercial sector is developing simulations to help model both system qualities and integration capabilities. These synthetic integration environments show both internal system integration capabilities and external interoperability. Numrich notes that many of these companies are using these capabilities to build integration environments for the Defense Department.

This leads to the next simulation challenge: ensuring that the simulation and modeling capabilities for one service’s contractor-based system work with the same capabilities for another service’s system operated by a different company. Numrich states that the Army’s Future Combat Systems (FCS) program “is as purple a capability as the Army ever has required, because [the Army] will rely on other service capabilities to give it some of their C4ISR [command, control, communications, computers, intelligence, surveillance and reconnaissance] capabilities, particularly the ISR capabilities.”

The DMSO must look beyond where it has gone so far with solely defense capabilities, Numrich offers, and instead must look at “the integration of integration capabilities.” And, standards lie at the heart of this evolution.

To examine how sensors function in diverse environments, for example, planners must create a number of different kinds of battlespaces. Achieving these goals probably will entail establishing private-public partnerships, Numrich offers. Some of these will be augmented with virtual reality and gaming capabilities.

“We start building architectures for our systems, and we rapidly end up with 1,200 pages of interface specifications,” Numrich charges. “I don’t think you make sense out of that. We are going to have to learn to put our hands around things and model them.”

Numrich adds that, as if all of these efforts were not complicated enough, now the department is moving full speed into the Global Information Grid, or GIG. The GIG will bring with it a whole new set of capabilities that will change the information infrastructure paradigm. With modeling and simulation a part of that infrastructure, the DMSO has become engaged with the Office of the Assistant Secretary of Defense for Networks and Information Integration (ASD NII) to implement modeling and simulation services on the GIG. The DMSO also is working with the ASD NII to ensure that the modeling and simulation naming conventions are aligned with those of operational C4ISR and the rest of the GIG community of interest.

Managing the GIG looms as a major integration challenge. Numrich allows that no real simulation or modeling capability currently exists to model the Internet. While some of the connection pieces can be modeled, no global model exists. Some experts estimate that it would take years to compute such a model.

One of the most exciting simulation developments in the next five to 10 years will be massively multiplayer online games, Numrich says. While the near future will see a “measurable capability” in this arena, the groundwork already has been laid. For example, the Army has designed and developed a massively multiplayer online game known as America’s Army. Not only has this online game proved to be an excellent—and inexpensive—recruiting tool, it also has helped some incoming cadets enter the U.S. Military Academy at West Point better equipped for the rigors of that academic program.

America’s Army not only teaches participants about an Army career, it also empowers the player with basic skills that can translate to the real world, Numrich offers. This is the same experience that has been built by flight simulator computer games, and the Army’s development of its own game established the “verifiable reality” that lies at the game’s core. The average age of the America’s Army game player is not in the teens but in the 20s.

The Defense Advanced Research Projects Agency (DARPA) is working on the DARPA Training Superiority program, or DARWARS. This effort aims to develop a massively multiplayer online environment that would be available constantly for a trainee. Numrich notes that several other types of games are online and running.

Many of these efforts represent the marriage of the simulation industry with the gaming industry. Now, computer engines are becoming powerful enough to make games less virtual and more real. Then, Numrich relates, the question will become, “What is it you can really do with it?”

Online gaming may lead to more esoteric applications for simulation. Several online games allow players to build virtual societies replete with economic rules. Future applications may permit users to model how another culture behaves under varying conditions. This could allow analysts to model politically sensitive situations.

Computational systems that are faster and less expensive also will open the door to multimedia capabilities that offer input to more human senses than just sight and hearing. “You are going to have this type of interactive capability,” Numrich says. “I don’t know that we’ll reach the [Star Trek] holodeck, but those of us who are in the simulation industry look at that as the ideal to aim for.”

Some future systems may be built on biological modes. These complex systems may be drawn from areas of biology based on a level of complexity theory that humans have not yet learned to use creatively.

Numrich reports that the DMSO is looking at Web services, which work well under a number of different circumstances. However, problems emerge because many users do not organize according to the same “mental model,” or ontology. This means of organizing, regarding and naming material is an important issue as the office moves to the Web, because search and recovery services work best with common ontologies, she says. “This is a part that nobody sees, but it is a part that is very important because you can’t get at it if you can’t name it right,” she emphasizes.

A partnership between government and industry is essential for building these ontologies, Numrich continues. “I’d ask industry to think carefully about proprietary and open source,” she says. “Unless we create the ontologies and the standards, when we go out into this new information paradigm on the Web, we will have just created a new way of not working together well.”

The rapid evolution of information technology is complicating future simulation efforts. Sometimes the rapid advances of these technologies tempt planners into moving forward too quickly “without shoring up our past,” Numrich suggests. However, this rapid evolution also is the key enabler of advanced simulation. “It is a two-edged sword,” she says.

For example, the DMSO has been testing Web-based technology for the past two years. But, Numrich reports that it is “a couple of orders of magnitude” too slow currently to do what the DMSO needs. However, she expects Web technology to be able to measure up soon. “It has been giving us the capability, and it has been giving us fits,” she states.


Web Resources
DMSO SEDRIS program:
America’s Army: