Advanced Decision Making Support using Intelligent Agent Technology

Abstract

DSS applications have demonstrated that intelligent agents can reduce complexity and produce new information or knowledge through integration of data sources, particularly in distributed systems. Capabilities beyond those traditionally associated with DSSs are required for application to autonomous robotic systems. The DSS must be able to deliver inference knowledge to allow the robotic system to make decisions autonomously and then perform some decision-making tasks without human intervention. This paper presents such an extension of research by presenting an intelligent-agent-enhanced DSS that enables autonomous decision making using inference knowledge in a high risk, time-pressured, dynamic environment in which communication with a human decision maker may be limited or unavailable. The approach allows dynamic problem structuring and explanation for policy prescription purposes rather than prediction through a static structure. Research results suggest that this type of enhancement has the potential to enable real-time decision support in complex and dynamic environments, including situations that benefit from a level of system autonomy.