Editorial of Special Issue on Distributed Intelligent Systems

Systems are becoming more and more complex. The properties of distribution are more and more important in systems for their identification, description, management, controlling, maintenance, knowledge, and intelligence. Applications of Distributed Intelligent Systems are emerging in more and more fields of industry. A distributed system is composed of many computers and devices interconnected with networks which cooperate and coordinate to accomplish one common task. Distributed intelligent systems are intelligent systems built on a distributed system. They are based on the use of cooperative agents and organized in hardware or software components. In the system, each agent independently handles a small set of specialized tasks and cooperates to achieve the system-level goals and a high degree of flexibility. Distributed intelligent systems are also simulations of people societies or virtual societies. To simulate real societies, we need to understand the nature of our societies. A well-organized society should encourage the members of the society to contribute. With the same requirement, distributed intelligent systems as virtual societies should also create a healthy platform or environment for virtual participants to contribute and work in the virtual societies. We may view agents as virtual persons in intelligent systems. Distributed intelligent systems are aiming to obtain collective intelligence by organizing a large quantity of simple agents. This special issue concentrates on the issues of intelligence and its distributions, management, controlling and maintenance of systems. I believe all the papers in this issue could provide the readers with new and broader views of distributed intelligent systems. The first paper by Benaskeur et al. reports on the project initiated by Defense R&D Canada-Valcartier that is intended as a vehicle to assess holonic control as a means of improving tactical sensor management for distributed military surveillance operations. Three levels of sensor management are considered: sensor, platform, and group. The proposed design is used to develop a simulation using a military scenario in which the holonic control system is employed in the sensor management role. Boloni et al. presents an agent-based coalition formation approach for disaster response applications. We assume that agents are operating in a dynamic and dangerous environment, and they need to form convoys to traverse unsafe areas. We introduce a commitment-based convoy model, where the commitments are negotiated between the participant agents. We show that this leads to a complex multi-issue negotiation, with two spatial and two temporal components. We propose an approach for reducing the negotiation …