Abstract
This paper discusses the necessity of studying new weapons systems with a comprehensive battlefield simulation. In the past, evaluations have been done with single event operations research models. These are useful, but today's systems are more interdependent than ever before. It is necessary to consider a system's contribution to the outcome of the battle and the availability of its support systems in carrying out its mission. A weapon may perform its designated function flawlessly but contribute little toward winning the war. A simulation which reproduces the entire environment in which a system will operate is a useful tool for answering complex questions about the utility of that system.
I. Introduction
The Combined Arms Systems Engineering (CASE) Center was started by General Dynamics to provide systems analysts and weapon systems designers with a tool to evaluate the performance and contributions of new weapons systems to a combat scenario. Proposed systems can be evaluated by an analytical computer simulation of the entire theater of combat (Fig. 1). These studies can uncover possible improvements and weaknesses of the system while it is still in the conceptual phase of development. This can save millions of dollars in development costs compared to making the changes once in the prototype phase, during full scale development, or after the asset has been fielded.
CASE synthesizes concepts from two communities within the simulation field. The first is the Operations Research community which for years has developed detailed analytical models of individual events found in the war fighting environment. Some models evaluate the effectiveness of weapons against ground targets such as tanks, trucks, bridges, and buildings. These have been used to study new air-to-surface missiles, bombs, cluster munitions, artillery warheads, and multiple launch rocket systems. A dogfight model was developed to determine the number and type of air-to- air weapons used in a fighter engagement and the number of aircraft killed on each side. The Terminal Surface-to-Air Missile Model (TSAM) analyzes the effectiveness of SAMs against maneuvering aircraft. There are numerous other models of this type and scope in the operations research inventory. They have typically provided excellent analysis of a single event, such as an engagement, given that the assets are able to get into the situation. They do not represent the effects of the support systems, logistics, air base runways, maintenance, and more importantly command and control functions, required to field the systems being modeled.
The second community integral to CASE is that of command training simulations. These are usually broad, theater level simulations that are used to stimulate commanders and give them the opportunity to test the tactics, doctrine, and strategy they plan to use in war time. Simulations like this usually include all assets that are significant to the outcome of a battle. The simulation has the depth necessary to provide commanders with reasonable feedback from their decisions, but the software is not meant to provide the level of detail found in an operations research model.
Both of these communities have been hindered in the past by the ability of computers to run their software in a reasonable amount of time. This has prevented them from expanding the scope and capabilities of the various models. The developers have also usually been pioneers, creating a simulation of something that has never been represented in software before. Therefore, they have not had the benefit of building on existing models and tools.
II. Air Land Battle Assessment Model
With the recent advances in computer technology, it is now possible to marry these two communities in a single simulation. This is what the Air Land Battle Assessment Model (ALBAM) is designed to be. It provides a single simulation with the operational breadth of command training simulations and the analytical depth of operations research models. This enables us to evaluate the performance and contribution of a new weapon system in the environment in which it will operate. System performance in this environment is a much more accurate measure of its worth than modeling it in a typical operations research program.
On the surface, ALBAM looks very much like a command training simulation. There are nodes in which military commanders make employment decisions based on the way they see the battle unfolding. These participants have three interfaces with the simulation software. The first is a graphics workstation which displays the terrain, road network, and tactical maps of the area of combat. Over this are laid the unit symbols which show the type, location, and formation of a commander's forces, as well as those of enemy units which have been detected by his sensors or those that report to him (Fig.2). More detailed information can be found on 23 different tabular information displays. These show information such as equipment compositions, unit capabilities, aircraft availability, and communication reporting networks (Fig. 3). Based on this information about the battlefield, a commander decides how to employ his forces. These decisions are injected into the software through input terminals which deliver orders to specific units.
The software is made up of several sections (Fig. 4). The air model controls functions such as the launching of aircraft from air bases, their navigation to an ordered location, and the performance of their mission. If the mission is an air-to-surface engagement, the aircraft will perform the delivery maneuver and the attrition results will be calculated by the high fidelity model imbedded in the ALBAM software (Fig. 5). The equipment killed will then be removed from the target unit; or fixed targets will be partially destroyed (e.g. 70% damaged) as a function of ordnance and structural integrity of the target. If the aircraft is on a surveillance mission it may take up station in a race track orbit while its sensors collect information about enemy forces. Units detected are reported to the command nodes via the communication reporting network. Other aircraft may perform air-to-air refueling missions allowing strike aircraft to reach deeper into enemy territory and surveillance aircraft to remain on station for longer periods of time. Aircraft also engage in air-to-air combat where the results are determined by response surfaces generated from the dogfight model described earlier.
All aircraft are subject to surface-to-air attack by missiles and anti- aircraft artillery (AAA). The results of a SAM engagement will be calculated using software which represents the TSAM model while AAA uses its own detailed model. Destroyed aircraft and their weapons are removed from the game as a result of these calculations. In this way the effect of SAM fire on air strike successes or aircraft surveillance capabilities can be measured.
As all of this is taking place ground forces are maneuvering across terrain and down road networks to carry out their assignments. Digital Terrain Elevation Data creates a landscape which affects unit movement and operations just as it would in real life. The shortest path between two points on the road network is calculated using a variant of Dijkstra's algorithm. Should they come into direct fire situations with the enemy the results of these are calculated using the Epstein Equations. Developed by Joshua Epstein, this methodology calculates equipment attrition and force movement on both sides in a dynamic environment. Some units may be able to bring indirect artillery or rocket fire against the enemy. These engagements are calculated using the detailed indirect fire methodologies.
As units move and engage in combat they consume their supplies. Logistics units automatically send convoys to resupply the ammunition and fuel being consumed. Combat engineering units move about the battlefield destroying bridges, emplacing minefields, and constructing barriers. These obstacles delay moving units and make them more vulnerable to air and ground attacks.
Helicopters and fixed wing aircraft can airlift personnel and equipment to any location on the battlefield. This is particularly useful in simulating special forces operations. SAM fire can destroy these aircraft and the assets being transported, thus impacting the success of the mission. '
A commander�s perception of the enemy is entirely dependent on the performance of the sensors that report to him. The tasking of these sensors, the intelligence reporting network, and the various delays associated with moving information through the network affect the construction of this picture. ALBAM simulates electro-optical, infrared, and active and passive radar sensors. The performance of these is dependent upon,
Though ALBAM represents nearly all significant assets present in battle, the developers of a new system often have their own detailed simulation of the asset. When studying it, they may want to include this in the exercise. Rather than reproducing this software inside ALBAM, we have developed interfaces that allow us to integrate the two into a single system. Interfaces can also allow ALBAM to stimulate real military equipment. This is particularly useful with intelligence, communication, and command and control shelters. The shelter can be stimulated in such a way as to make it operate just as it would in an actual combat situation (Fig. 6).
As the war develops, it is essential that the results be captured for analysis. Data extraction modules collect the outcome of every engagement, the performance of the system being studied, its contribution to other forces in the exercise, and other measures of effectiveness. We can also quantify the synergisms taking place between different systems on the battlefield (Fig. 7). It is important to capture the decisions and intentions of the human commanders participating in the exercise. This is done through a combination of player survey sheets and software. All of this data forms the analytical basis for the reports generated from a study. This allows us to measure the effects of such things as destroying the new asset's support structure or jamming its sensors.
Studies usually involve several replications of a simulation, both with and without the asset being analyzed. Cases "without" provide a baseline from which to measure the contribution of the asset. Each exercise may simulate several hours or several days of combat. Depending upon the study, these may be run at real time or at speeds from two to ten times faster than real time. Since ALBAM involves command decisions, it is usually necessary to run at real time. However, for periods when intense command interaction is not necessary it is possible to increase the speed. This shortens the duration and lowers the cost of the study. Increased speed is possible for night time periods when the operations can be planned ahead of time, entered into the simulation, and left to execute without human interface or decisions.
Since several cases are being run it is necessary to block out the effects of learning by the commanders. A possible exercise schedule, man loading, and participant design is illustrated in Figure 8.
Periods are scheduled between exercises to allow for data analysis and storage, database and software preparation, system maintenance, and participant team training.
III. Applications
The Combined Arms Systems Engineering Center has used ALBAM to analyze several different weapons concepts. These include the contribution of a new airborne sensor to the US capability to interdict enemy forces with air and artillery assets. We are able to measure the performance of the sensor as well as the utility of the detections it provides to artillery and air units. It is possible to determine whether the asset provides useful information that is not currently available and thus fills a hole in our defense structure.
We have studied the contribution of cruise missiles used in conventional combat situations. The effects of the missiles themselves against specified targets was measured. But, more significantly, we noted an increase in the effectiveness of the air forces as a result of their ability to reallocate aircraft that would have been assigned to the cruise missiles' targets.
An improved communication system was studied to determine the effects of providing commanders with more timely information about battlefield situations. Three different scenarios were run:
IV. Conclusion
It is absolutely necessary to study new systems in the context of the environment in which they will participate. We must measure the interdependencies that are becoming a bigger and bigger part of a system's contributions. Total battlefield simulations provide a mechanism for these types of evaluations.
We expect analytical battlefield simulations to become an important part of weapon systems design. They have the potential to save the government millions of dollars in development and procurement costs. Though trial and error is very interesting, it is also a very expensive method for developing new capabilities. Moving these experiments into a computer simulation will allow us to eliminate poor designs long before money is spent building a prototype. Just as operations research has improved the quality of products and training simulations have improved the capabilities of our commanders, these new analytical simulations will allow us to develop superior products in less time and at lower costs in the future.
Acknowledgement : The author thanks many present and former members of the Combined Arms Systems Engineering Center for their work, which forms the basis of this article. Special thanks to Gary Christopher for providing the data analysis charts and Kevin Stafford and Stan Hall for editorial comments. The work of the following people is also described in the article: P. Wu, M.G. Meeks, S.L. Tucker, J.K. Pickering, R.C. Toifel, K.R. Stafford, K.O. Tucker, J.K. Goodridge, W.E. Carroll, J.F. Zelenevitz, C.N. Dollin, J.R. Runyon, J.R. Hester, M.S. Giles, D.L. Finley, J.D. Pritchett, E.H. Nathman, B.C. Madsen, D.C. Meyers, V.T. Le, K.E. Tabor, S.G. Ferreira, M.H. Piazza, G.L. Christopher, D.R. Cooper, T.J. McKnight, J.L. Brunick, R.M. Forman, and J.S. Hecht.
Author Summary :
ROGER D. SMITH is a Senior Operations Analyst at the Combined Arms Systems Engineering Center in Fort Worth, Texas. For the past five years he has been developing analytical battlefield simulation software and operations research models. He received his Master of Science degree in Statistics from Texas Tech University in 1985 and his Bachelor of Science degree in Applied Mathematics from the University of Southern Colorado in 1978. Currently he is working toward an MBA at Texas Christian University. He speaks and teaches regularly on mathematics, simulation, and business topics.
Bibliography :
1. Epstein, Joshua M., "The Calculus of Conventional War: Dynamic Analysis
Without Lanchester Theory", The Brookings Institute, 1985.
2. "Joint Warrior Simulation: Air/Ground Combat Model Family
Documentation", Third Edition, 1985.
3. "JASMEAN User's Manual", Version 3.1, General Dynamics, 1989.
4. "Air Land Battle Assessment Model Methodology Manual", General
Dynamics, 1990.
5. "Conventional Combat 2002: Battle Management/C3I Technologies proof-
of-Concept Evaluation Study", Combined Arms Systems Engineering Center,
General Dynamics, 1990.
6. "Terminal Surface-to-Air Missile Model User's Guide", General Dynamics,
1990.
7. "Derivation of JMEM/AS Open-End Methods", 61 JTCG/ME-3-7, Joint
Technical Coordinating Group for Munition Effectiveness, 1983.
8. Tarjan, Robert Endre, Data Structures and Network Algorithms,
Society of Industrial and Applied Mathematics, Philadelphia, PA, 1983.
9. Fredman, Michael L. and Tarjan Robert Endre, "Fibonacci Heaps and Their
Uses in Improved Network Optimization Algorithms", Proceedings of the
IEEE, 1984.
10. "Programmable Calculator manual for Evaluating Effectiveness of
Nonnuclear Surface-to-Surface Indirect-Fire Weapons Against Area targets",
61-JTCG/ME-77-14, Joint Technical Coordinating Group for Munition
Effectiveness, 1978.
11. Toms, R. M., "Adaptive Line of Sight Algorithms", Lawrence Livermore
National Laboratories, 1987.
12. Olson, W. K., McCoy, D. H., Champion, D., Washer, R., "Tactical Terrain
Intervisibility Classification Study", U.S. Army TRADOC Analysis Command,
1986.
13. Epstein, Joshua M., Conventional Force Reductions: A Dynamic
Assessment, The Brookings Institution, Washington D. C., 1990.
14. Allen, Thomas B., War Games, Berkley Books, New York, 1987.
15. D.Y. Harwell, editor, "Battlefield Integration Center Simulation: User's
Guide, Version 1.6.1", 1991.
