Distributed intelligent systems: What makes them "intelligent"

Three core requirements are discussed for the implementation of distributed intelligent systems: a peer-to-peer communications infrastructure, a multi-agent architecture, and a means to provide distributed logistics for the management and self-organization of system resources. We distinguish between a distributed system and a distributed intelligent system, and the inherent advantages that they have over centralized methodologies. By incorporating intelligence in a distributed system, it is capable of cooperation, collaboration and coordination, features that have cardinal importance for highly configurable networks such as ad-hoc systems. We highlight research being conducted by Simon Fraser University and Intelligent Robotics Corporation to design and implement wireless distributed intelligent systems.