Autonomous Multi-Agent Decision and Control

Period of Performance: 11/18/1998 - 11/18/2000


Phase 2 SBIR

Recipient Firm

American GNC Corp.
888 Easy Street
Simi Valley, CA 93065
Principal Investigator


This Phase II project is intended to develop Autonomous Multi-Agent Decision and Control techniques and demonstration systems. The methodologies and techniques for multi-agent system modeling and analysis, distributed information fusion and optimal multi-agent decision-making and hybrid control will be conducted and developed. The emphasis will be on research in the applicability of these techniques to battlefields. In order to demonstrate and evaluate state-of-the- art multi-agent decision and control techniques, two practical simulation systems: Multi-Agent Battlefield Simulation and Intelligent Minefield (IMF) Simulation will be fully developed following the Phase I work. The simulation systems possess the capabilities of graphical display of environmental fields and decision-making and communication processes for easy and realistic visualization. They will be generalized to be an excellent tool for development and evaluation of multi-agent decision-making and hybrid control techniques as well as practical autonomous multi-agent systems. Also, key issues related to real-time implementation and applications will be addressed. The developed techniques and obtained results in this project will be transferred to our commercial product: Multi-Agent Decision and Control toolbox, which has wide application potential in civilian and military areas, such as the National Automated Highway System. BENEFITS: The Phase II results are applicable to manufacturing, precision machine tools, process control, engine control and automation applications, including automobile and commercial aircraft manufacturing, robotics, flight controls, smart highway systems, etc. The defense applications arise in all areas of smart weapons, robotics, defense manufacturing, and command and control. The impact of the technology is two-fold: increased control performance and accuracy through improved software and reduced operating and support costs through automation, fault tolerance and on-line adaptation.