Network-centric IED detection planning

Abstract

As methods for detecting improvised explosive devices (IEDs) continue to diversify, it becomes increasingly important to establish a framework for coordinating distributed IED monitoring resources to best protect a designated area. The purpose of this paper is to establish the beginnings of such a framework in a distributed plan execution context. The first contribution of this paper is defining an automated planning domain for distributed IED detection. In doing so, we investigate approaches for coordinating distributed plan execution resources. Whereas many existing multi-agent system (MAS) frameworks abstract network information from agent decision-making processes, we instead propose that MAS frameworks consider network properties to improve effectiveness. The second contribution of the paper is the description of several types of network-aware planning, execution, and monitoring agents and a comparison of their performance and effectiveness in an IED monitoring scenario. The results of this research indicate that network-awareness improves distributed plan execution (e.g., IED monitoring). Furthermore, we find that an agent’s benefit from reasoning about network properties is directly affected by the amount of dynamism in the network.