11. Intelligent Command Aids (UK)
Description: Project agreement is to investigate potential payoff associated with incorporating emerging fuzzy logic techniques into a large scale battlefield decision making simulation.
Justification: Armies throughout the world continue to expend considerable resources in the pursuit of developing intelligent command aids, particularly in the area of simulation and computer generated forces. A common problem is that of representing the command and control process. It is far too expensive to have human controllers "command" the computer generated forces. Current approaches have usually used large rule-based systems to construct "command agents" that attempt to model individual decision making entities. However, the deterministic behavior introduced by the rule set and inference engine does not realistically reflect the need to act or vague (or fuzzy) knowledge characteristics of a real-world command and control problem. Fuzzy logic techniques and fuzzy inference engines are an approach that enhance current intelligent command aids by overcoming this problem and may also provide proven methodologies for the construction of more realistic and effective future systems.
Implementation: The GeKnoFlexE system developed by the UK (Ft. Halstead), which simulates battlefield decision making, will be used as the testbed system. The current conventional rule-based inference structure will be "fuzzified" by augmenting and/or replacing it with fuzzy rules and a fuzzy inference mechanism. Since the current GeKnoFlexE system is completely non-fuzzy, direct comparisons of complexity, behavior, and other performance parameters will be possible.
Dr. Rodney Smith
Army Materiel Command
5001 Eisenhower Blvd
Alexandria, VA 22333-0001
Mr. Stephen Cohn
Army Research Laboratory
2800 Powder Mill Road
Adelphi, MD 20783-1197