Intelligent Agent Research

Today, large amounts of critical information, scattered around the battlespace, cannot be located and processed fast enough for timely decision making. The result is poor situational awareness because commanders lack the facts they need. Intelligent software agents are now being developed that will collect, appraise, and deliver pertinent information to the right component at the right time. Yet these agents lack a common framework to interoperate in battlefield applications. The challenge is to research a common framework that permits legacy, current, and future agents to function together at the tactical battle command level, communicating across a variety of networks, to provide information to decision makers in real time. GITI will research the variety of currently available agent technologies and frameworks. We will determine their strengths and weaknesses and document the requirements of a common agent development environment. We will use those requirements with our extensive experience in agent technology to design an Agent Factory. This Factory will allow non-engineers to create agents that will interoperate with each other, other agent frameworks, and legacy systems. This will allow tactical personnel to tailor agents to react quickly to changing situations such as hostile threats, movement of coalition forces, or modified operation plans.