Major John R. "Buck" Surdu
Buck Surdu is a U.S. Army Major attending A&M in pursuit of a Ph.D. in Computer Science. He received a B.S. degree in Computer Science from the United States Military Academy (West Point) in 1985, a Master's in Business Administration in 1991 from Columbus State University, and a M.S. in Computer Science from Florida State University in 1995. His areas of interest are simulation technologies and artificial intelligence with particular focus on the synergies created by combining these two fields. Upon graduation, Buck Surdu will be assigned to the faculty at West Point in a research/teaching position in the Information Technology and Operations Center.
A large number of tasks performed by commanders and staff can be facilitated during operations by the application of simulation technologies. Traditionally the focus of simulations within the Department of Defense has been on analysis and training. Simulations designed to facilitate course of action development and analysis, rehearsal, and operations monitoring will enhance the effectiveness of staff and commanders. Currently there are no operationally focused simulations built specifically for use during operations.
The Army Modeling and Simulation Office (AMSO) recognized the importance of simulation in command and control and identified five voids in current modeling and simulation technology for the Army in the future. According to their analysis, the first of these voids, Cognitive Modeling, includes automated decision aids, course of action tools, and tactical information aids. The methodology proposed in this research intends to address these three issues.
This methodology involves a specifically designed, operationally focused simulation used for course of action analysis and design. Once the operation commences, this simulation "plays back" the plan in near real time in parallel with the real operations. Between the planned operation (simulation) and the real operation lies a hierarchy of rational agents. These agents, termed Operations Monitors (OMs), compare the plan with the actual operations. When the OMs note "significant" differences, they can launch other agents or tools to analyze these differences. If the OMs determine that the differences significantly impact the probability of success of the operation, they take advisory actions. Other OMs are also involved with resynchronizing the simulation to the real world and updating simulation parameters to make it a better model of the real world.
OMs can use any appropriate reasoning mechanism, but the prototype system uses simple fuzzy rule bases. In order to make the analysis of a very complex and semantically charged domain tractable, each OM focuses on only a small portion of the domain, such as strengths of friendly units.
This presentation will provide a brief overview of simulation technology. Then it will describe the proposed methodology, paying particular attention to the way that the hierarchy of monitors expands and contracts dynamically. The discussion of why and how OMs resynchronize the simulation will also be discussed. Finally a brief description of the prototype system will be presented along with some preliminary results.
Everyone is invited and welcome to attend the seminars in this series.