Intelligent Learning Agent System for Advanced Targeting Decision-Aiding

The digitization of the tactical military environment with its concomitant increase in accessible information provides an opportunity to enhance decision-aiding capabilities of advanced Navy targeting systems. An approach employing intelligent software agents can facilitate enhanced decision-aiding by automating: the management and dissemination of relevant targeting information and intelligence data; the coordination and communication required for supporting distributed multi-site/multi-user operations; and the interface to tactical C4I systems. Additionally, the learning and tutoring aspects of agent technology can address user customization and on-line in-context coaching. Here, we propose an intelligent learning agent system for advanced targeting decision-aiding. Our system incorporates an off-line learning agent for learning operator responses to given tactical state information and event changes from prior event histories, and an on-line agent that incorporates an inference engine to map current event/state information into actions (in the autonomous decision-making mode) or suggested actions (in the tutoring/coaching mode). Our key innovation is the incorporation of several synergistic technologies: distributed processing and inter-agent communication capabilities, embedded knowledge, inferencing power, and on-line learning and customization of software agents. We see considerable potential for this approach in enhancing Navy advanced targeting decision-aiding systems and addressing future digitized battlespace requirements.