WSC 2004 Final Abstracts
Monday 10:30:00 AM 12:00:00 PM
Panel Session: Military Senior Leadership
Chair: Gregory McIntyre (Applied Research Associates)
View From the Top: Military Challenges for the Simulation Community
Gregory A. McIntyre (Applied Research Associates) and Raymond R. Hill (Wright State University)
The Department of Defense (DoD) has become increas-ingly reliant on models and in particular on simulation models. The military-defense establishment and its com-bat-preparation orientation is one of the most complex systems in existence, particularly in the extremely dynamic modern world. Simulation models form the basis for analy-ses spanning issues ranging from force structuring to ac-quisition prioritization. These analyses, and the subse-quent decisions they support, mold and shape the DoD thereby influencing the posture of the US defense estab-lishment. This panel brings together a set of the military's influential decision makers directly involved in the development and use of simulation models. The panel will discuss their cur-rent and anticipated needs for the future of simulation and pose the challenges the simulation community must meet to ensure those future needs are met.
Monday 1:30:00 PM 3:00:00 PM
Human Systems Modeling I
Chair: Janet Miller (Air Force Research Laboratory)
CHI Systems, under contract to the U. S. Army Research Institute, is developing an immersive training system, called Virtual Environment Cultural Training for Operational Readiness (VECTOR), which applies highly experiential, scenario-based virtual environments to training in cultural familiarization. To produce an interactive, realistic training environment, the simulation must incorporate synthetic ac-tors, or non-player characters (NPC’s), that are capable of evaluating and responding to the cultural propriety of trainee’s actions. The main focus of the paper is on ex-plaining how iGEN™ cognitive modeling architecture is being used to create executable cognitive models and emo-tion models which inform and constrain the overall reac-tions and behaviors of NPC’s toward the trainee. In addition to influencing the behavior of the active NPCs, the emotion models constrain interactions with NPC’s encoun-tered later in a scenario. In this way, the training system provides a means of modeling the overall cumulative emo-tional state of the simulated population.
Common Problems and Helpful Hints to Solve
Them: Lessons Learned in Integrating Cognitive Models in Large-Scale Simulation
Karen A. Harper and Greg L. Zacharias (Charles River Analytics, Inc.)
The application of M&S simulation technologies to advanced analysis and training functions throughout the DoD has led to an increasing need for higher fidelity representations of human decision-making behavior than is currently available in most military simulation behavior engines. The appropriate path to meet this need is to incorporate cognitive models from the Human Behavior Representation (HBR) community that provide psychologically-rooted representations of decision-making behavior and performance. There are significant challenges associated with the integration of these models within complex simulation environments, however. Here, we attempt to identify some of these challenges and provide design strategies to overcome them. Specifically, we provide strategies for selecting appropriate modeling resolution for specific applications, dynamically managing the resolution of those models throughout a simulation run, and dealing with the general mismatch of sensor and control data between simulation environments and HBR models.
Use of Intelligent Controller Nodes to Augment Human Role-Players in Synthetic Battlespace Exercises
Gary Buker (Raytheon) and Brian C. Flis (Air Force Research Laboratory)
Distributed exercises, which combine live, virtual, and constructive entities, are increasingly used for training, readiness exercises, and CONOPS Analysis. These exer-cises require the support of human controllers who monitor and control simulated entities to maintain realism of the exercise. The high cost in time, money, and personnel for supporting large exercises makes it important to reduce the number of role players required. In addition, the large vol-ume of information flowing from the synthetic battlespace and the narrow window for response dictate a need for mechanisms to assist the human role players who control the simulated entities. The Role Player Intelligent Control-ler Node (RPICN) uses intelligent agents to monitor and interact with the simulation environment. The agents can recognize significant events, make recommendations, and perform routine actions for the role player. By increasing situational awareness and automating routine tasks, the RPICN improves the efficiency of the human controllers.
Applying a Cognitive Architecture to Control of Virtual Non-Player Characters
Christopher McCollum, Charles Barba, Thomas Santarelli, and John Deaton (CHI Systems, Inc.)
Monday 3:30:00 PM 5:00:00 PM
Human Systems Modeling II
Chair: J. Miller (Air Force Institute of Technology)
Approaches for Modeling Individuals Within Organizational Simulations
Eva Hudlicka (Psychometrix Associates Inc.) and Greg L. Zacharias (Charles River Analytics Inc.)
The human behavior modeling community has traditionally been divided into those addressing individual behavior models, and those addressing organizational and team models. It is clear that these extremes do not reflect the complex reality of the mutually-constraining interactions between an individual and his/her organizational environ-ment. In this paper we argue that realistic models of or-ganizations may require not only models of individual de-cision-makers, but also explicit models of a variety of individual differences influencing their decision-making and behavior (e.g., cognitive styles, personality traits, and affective states). Following a brief review of individual differences and cognitive architectures research, we de-scribe two alternative approaches to modeling the individ-ual within an organizational simulation: a cognitive archi-tecture and a profile-based social network. We illustrate each approach with concrete examples from existing proto-types.
Exploring the Constraints of Human Behavior Representation
John C. Giordano, Paul F. Reynolds Jr., and David C. Brogan (University of Virginia)
Human behavior representation (HBR) is an elusive, yet critical goal for many in the simulation community. Requirement specifications related to HBR often exceed cur-rent capabilities. There exist a number of tools, techniques and frameworks to model and simulate HBR, but they are constrained and do not generalize well. Even with a vibrant research community, certain HBR characteristics remain beyond our grasp, unless some unforeseen disruptive technologies emerge. We survey the state of the practice for HBR, discuss ongoing research, and identify what appear to be insurmountable challenges. Along with exposing the essential characteristics of HBR and their current level of maturity, we propose a generational framework for considering HBR capabilities. While a number of HBR issues have been addressed in the literature, there is no published discussion explicitly detailing its constraints and limitations.
Validation of the Enlisted Grade Model Gradebreaks
Andrew O. Hall (Army )
This paper describes the validation of the Enlisted Grade model gradebreaks and describes a current application of simulation in operations research. The Enlisted Grade model is part of a suite of models used by Army strength analysts to forecast active Army strength to develop the Active Army Military Manpower Program and for use in the President’s Budget and the Program Objective Memorandum. The Enlisted Grade model gradebreaks were tested by predictive validation, comparison to the Military Occupational Specialty Level System and face validity before acceptance. The validation of the Enlisted Grade gradebreaks was the final milestone in the acceptance of the Enlisted Grade model.
Tuesday 8:30:00 AM 10:00:00 AM
Military Analysis Using Agent Models
Chair: Subhashini Ganapathy (Wright State University)
Agent - Based Model of Auftragstaktik: Self Organization in Command and Control of Future Combat Forces
Robert H. Kewley (Center for Army Analysis)
Agent-based modeling is a framework that allows the analysis of distributed command-by-influence using mis-sion-type orders known for over a century as Auftragstaktik in German Army manuals. A combat simulation with embedded decision agents analyzed this type of decentral-ized command and control. Local commanders relied on their improved situation awareness, mission goals, and constraints from higher commander to drive a small set of robust decision methods. In this sense, they self-organized, and an effective set of actions for mission ac-complishment emerged. The improvement was most dra-matic for more capable future combat forces. Agent-based modeling provides a laboratory for experiments in com-mand and control of future combat forces.
Simulated Annealing for Selection of Experimental Regions in Response Surface Methodology Applications
Jeffrey B. Schamburg (United States Military Academy) and Donald E. Brown (University of Virginia)
In this paper we describe a methodology that includes the complementary use of simulated annealing and response surface methodology (RSM). The methodology was de-veloped for analysis of simulations to help determine pro-cedures for the employment of superheterodyne surveil-lance receivers. In this methodology, we use simulated annealing to determine near optimal solutions and to help select an initial search region from which to begin experi-mentation and analysis. By using this technique, we are able to take the results of an otherwise obscure function, over a limited range of the variable values, and develop a simplified, more understandable model which closely represents the actual system over the limited solution space.
Assessing Obstacle Location Accuracy in the Remus Unmanned Underwater Vehicle
Timothy E. Allen (United States Navy) and Arnold H. Buss and Susan M. Sanchez (Naval Postgraduate School)
Navy personnel use the REMUS unmanned underwater vehicle to search for submerged objects. Navigation inac-curacies lead to errors in predicting the location of objects and thus increase post-mission search times for explosive ordnance disposal teams. This paper explores components of navigation inaccuracy using discrete event simulation to model the vehicle’s navigation system and operational per-formance. The simulation generates data used, in turn, to build statistical models of the probability of detection, the mean location offset given that detection occurs, and the location error distribution. Together, these three models enable operators to explore the impact of various inputs prior to programming the vehicle, thus allowing them to choose combinations of vehicle parameters that reduce the offset error between the reported and actual locations.
Tuesday 10:30:00 AM 12:00:00 PM
Military Modeling Analysis Methods
Chair: Joshua McGee (University of Arkansas)
Poly-Functional Intelligent Agents for Computer Generated Forces
Matteo Brandolini and Attilio Rocca (BRB Studio), Agostino G. Bruzzone (McLeod Institute of Simulation Science - DIP Genoa University) and Chiara Briano and Petranka Petrova (Liophant Simulation)
The authors present the requirement definition and methodological approach for developing a new generation of Computer Generated Forces (CGF) based on Intelligent Agents. The analysis is based on the author initiative (PIOVRA) devoted to the development of High Level Architecture (HLA) components to be integrated for training, planning and operative support. These agents require to act as smart entities in the scenarios reproducing both enemies and/or civil units; their model includes characteristics that are typical of the psychological aspects affecting troops and soldiers on the battlefields as well as operative basic tactics. The authors avoid to create intelligent units in a wide sense due to the fact that Artificial Intelligence (AI) technology is quite far to the creation of an "intelligent virtual strategists or commander"; however their entities are expected to face set of complex situations with realistic cooperative and competitive understanding and reproducing real reactions to the boundary conditions.
A Generalized Multiple Response Surface Methodology for Complex Computer Simulation Applications
Jeffrey B. Schamburg (United States Military Academy) and Donald E. Brown (University of Virginia)
This work provides a generalization of the traditional re-sponse surface methodology (RSM) that can be applied to complex, multi-objective simulation studies. These prob-lems involve a larger number of input variables, multiple measures of performance, and complex systems relation-ships. This multiple RSM approach capitalizes on the un-derlying learning philosophy of the traditional RSM while benefiting from other knowledge discovery concepts and data mining techniques. Furthermore it does not require the restrictive assumptions of the traditional RSM nor does it restrict the analyst to the traditional RSM techniques. Based on a variation of [Brown, 2004] and [Schamburg, 2004], a brief description of the generalized approach is provided. Then, the multiple response techniques are shown through an example application.
Using Simulated Data in Support of Research on Regression Analysis
Christopher Michael Hill (Center for Army Analysis) and Linda C Malone (University of Central Florida)
Using simulated data to develop and study diagnostic tools for data analysis is very beneficial. The user can gain insight about what happens when assumptions are violated since the true model is known. However, care must be taken to be sure that the simulated data is a reasonable representation of what one would usually expect in the real world. This paper discusses the construction of simulated data sets and provides specific examples using linear and logistic regression analysis. It also addresses the execution of simulation based data studies following data construction.
Tuesday 1:30:00 PM 3:00:00 PM
Logistics and Urban Modeling
Chair: Jeffrey Schamburg (United States Military Academy)
Military supply chains encompass a complicated network of customers and suppliers, and deal with a wide variety of items. Demand inside the network is generated at the unit level at a specific base. The demand from the bases is ag-gregated to military service depots, which comprise the wholesale level in the network. The many layers of the supply chain often result in unnecessary cost and delay times, as well as low network reliability. Better integration between the multiple levels of the supply chain may be achieved through the effective utilization of transportation modes and criterion. In this paper, we present a simulation for quantifying the effect of transportation options (i.e. truckload shipping, less-than-truckload shipping, trans-shipments, and express air shipping) on shipping costs and operational availability.
An Urban Terrain Abstraction to Support Decisionmaking Using Recursive Simulation
John B. Gilmer, Jr. (Wilkes University)
Recursive simulation is the technique of having simulated decisionmakers themselves use simulation to inform their decisionmaking. Issues of efficiency require that the recursive runs, especially if extended over multiple levels to represent adversarial planning, be at a relatively simple and abstract level of detail. However, the nature of urban terrain is that it is dominated by particulars that drive up the level of detail and make automated decision representations difficult. The proposed terrain abstraction is intended to address these issues, and support low resolution embedded simulation used for recursive decision support for entities in a higher fidelity simulation.
Localization Using Discrete Event Simulation
Baybora Aksoy, Volkan Ustun, and Jeffrey S. Smith (Auburn University)
In this paper, we focus on the implementation of a localization algorithm for sensor networks using a discrete event simulation (DES) architecture. In this implementation, DES is used to calculate the distances between the sensors in terms of hop lengths and to implement a mass-spring optimization scheme. This implementation allows us to estimate the number of packets required during the localiza-tion process after the deployment of the sensors.
Simulating Transportation Practices in Multi-Indenture Multi-echelon (MIME) Systems
Joshua Burton McGee, Manuel D. Rossetti, and Scott J. Mason (University of Arkansas)
Tuesday 3:30:00 PM 5:00:00 PM
Simulation for Army Requirements
Chair: Raymond Hill (Wright State University)
The purpose of this research is to demonstrate the usefulness of integrating human-in-the-loop simulations and agent based modeling. The integration of a human-in-the-loop simulation with an agent based model can model in-formation technology systems. This integration allows analysts to exploit the strengths and advantages of each of these two model types. The integration and power of these models together diminishes each of the models own inherent disadvantages and limitations. This unique partnership between two distinct model types can tell analysts how well information technology systems provide users with information, data, and intelligence. This valuable insight about information systemsˇ¦ performance can be an indispensable aide to those interested in comparing, rating, and acquiring alternative information systems.
Simulation Modeling Requirements for Determining Soldier Tactical Mission System Effectiveness
Eric S. Tollefson, Michael J. Kwinn, Jr., Phillip G. Martin, Gregory L. Boylan, and Bobbie L. Foote (United States Military Academy)
In order to maintain an edge during this time of unprecedented technological growth, the Army must field Infantry soldier systems quickly; however, the cost of doing so without some assessment of utility is quite high. Therefore, the acquisition community must estimate the operational impact of proposed systems with an increasing degree of accuracy. For this, the Army has turned to combat simulations. However, the focus in the past has been on larger battlefield systems and unit-level analyses. Additionally, Infantry soldier models require unprecedented fidelity in terms of the soldier entity and his environment. As a result, the simulation representation of the individual soldier on the battlefield has not kept pace with other representations. In this paper, we discuss our identification of the unique simulation requirements for modeling the Infantry soldier as a system of systems in support of acquisition decision making.
Medium Caliber Cannon Lethality Study for Future and Current Infantry Fighting Vehicles
Russell Joseph Schott, Patrick M. Downes, Rocky Gay, Nate Whitten, James Paine, Michael Goddard, Michael Rybacki, and William Klimack (United States Military Academy)
According to the Army Chief of Staff, the Army’s infantry fighting vehicle, the Bradley Fighting Vehicle, will be in service until 2032. The Bradley Fighting Vehicle needs an improved medium caliber cannon to defeat the growing threats from improved light armored vehicles and hand held rocket propelled grenades. The Army can continue to keep the Bradley Fighting Vehicle in service by increasing the lethality of its weapon systems. We examine six medium caliber cannons and their impact on the battlefield. We also examine the use of new medium caliber air burst munitions. Combat modeling and simulation using the Joint Conflict and Tactical Simulation (JCATS) is used to predict the contributions of these new technologies to the infantry soldier. Multiple mission scenarios in different environments to include Baghdad urban combat are examined. The medium caliber cannon selected will be the final lethality enhancement for the Bradley.
Using Agent-based Modeling and Human-in-the-Loop Simulation to Analyze Army Acquisition Programs
Patrick M. Downes, Michael J. Kwinn and Jr. (United States Military Academy) and Donald E. Brown (University of Virginia)
Wednesday 8:30:00 AM 10:00:00 AM
Unmanned Aerial Combat Vehicles
Chair: Todd Hausman (Wright State University)
Recent airspace restrictions in Kabul have limited the potential capability of Tactical Unmanned Aerial Vehicles (TUAVs) within the area of operations of the Kabul Multinational Brigade. An experiment was conducted using the OneSAF Testbed Baseline and a range of virtual simulations to examine the impact of five different radar support options and three different information displays on the level of airspace situational awareness (SA) of the air traffic control officer. The SAGAT, SART and NASA-TLX techniques were used to determine differences in SA and workload. Simultaneous data capture through shared EXCEL workbooks and macros permitted near real time analysis. The Mann-Whitney U test, used due to the nature and limited size of the data sets, showed that any of the radars examined in this experiment would significantly enhance SA during TUAV operations in controlled airspace over Kabul.
Investigation of Error Rates When Controlling Multiple Uninhabited Combat Aerial Vehicles
Sasanka Prabhala and Jennie Gallimore (Wright State University)
As systems become more and more complex the use of automation tools becomes more important. Although automation is introduced to reduce human workload, im-prove situational awareness, and system reliability, in-creases in automation features also increase the overall complexity of the system. Despite the fact that research has been and is being conducted investigating the effects of automation on human performance, the results are often contradictory. This suggests the need for a universal way of presenting results so that trade-offs can be carried out between different studies. The purpose of this research was to investigate how a decision structure approach might be used as an aid for designers and researchers to conduct de-sign trade-offs when designing user interfaces for Unin-habited Combat Aerial Vehicles (UCAVs).
Joint Modeling and Analysis Using XMSF Web Services
Arnold H. Buss (MOVES Institute) and John Ruck (Rolands & Associates Corporation)
This paper describes the creation of a new analytical mod-eling capability by bringing together the Naval Simulation System (NSS) for sea strike and CombatXXI for littoral and land warfare modeling. The models are linked by Web services using principles from the Extensible Modeling and Simulation Framework (XMSF). This implementation is an examplar for a transformational framework for design, de-velopment, and integration of simulation models.
Measurement of Air Traffic Control Situational
Awareness Enhancement Through Radar Support Toward Operating Envelope Expansion
of an Unmanned Aerial Vehicle
James S. Denford (Directorate Land Synthetic Environments) and John A. Steele, Roger L. Roy, and Eugenia Kalantzis (Defence R&D Canada - Operational Research Division)
Wednesday 10:30:00 AM 12:00:00 PM
Advanced Concepts for Military Modeling
Chair: Sasanka Prabhala (Wright State University)
The British Army currently embraces a manoeuvrist style of command and hopes to gain further operational advantages from Network Enabled Capability (NEC) and effects-based planning (EBP). Many new communications equipments are being procured. Existing approaches to military simulation modeling have, for good reasons, not concentrated on command and control (C2), but this is now changing. The author proposes an approach to the modeling of military command systems based on Searle’s theory of speech acts, and suggests that it may have broader application than modeling C2 alone.
Using Morris' Randomized Oat Design as a Factor Screening Method for Developing Simulation Metamodels
Fasihul M. Alam (Brunel University) and Ken R. McNaught and Trevor J. Ringrose (Cranfield University, RMCS Shrivenham)
Simulation metamodels have been used for optimization, prediction, sensitivity analysis and understanding of complex, real-world systems. Since most simulation models contain a large number of input parameters, it is of great interest to determine the most important ones to include in a metamodel given a particular modeling context, i.e. given a particular set of questions which are to be addressed by the metamodel. This paper employs Morris’ randomized one-at-a-time (OAT) design as a factor screening method prior to developing a number of simulation metamodels. The approach is illustrated with reference to a stochastic combat simulation, called SIMBAT.
Developing Federation Object Models Using Ontologies
Tarun Rathnam and Christiaan J. J. Paredis (Georgia Institute of Technology)
The reuse of existing simulations in multiple federations is an important goal of distributed simulation frameworks. However, in order to reuse a federate, its simulation code often has to be modified so as to comply with the object and interaction representations defined in a corresponding Federation Object Model (FOM). Such modifications imply added time and effort, which diminishes the efficacy of reuse in federation development. In this paper, we present an ontology-based framework for modeling federates and supporting their reuse in multiple federations. Ontologies are used to specify the semantics of objects and interactions in federate domains in a formal, computer-sensible fashion. Using these formal semantics the relationships between federate simulation concepts are described in a reusable fashion. In doing so, a suitable federation representation for a set of related federate concepts and the required set of transformations between federate and federation representations are automatically derived.
Speech Acts of War
John David Salt (General Dynamics UK)