Emergent Intelligence in Large Scale Multi-Agent Systems

The authors with their development teams have designed and implemented multi-agent systems that exhibit emergent intelligence defined as the ability to find effective solutions to problems under conditions of uncertainty. One of these systems was designed specifically for road transportation logistics (referred to as the scheduler in the further text) and is described in some detail below. The paper describes a multi-agent system which is capable of achieving its goals under conditions of uncertainty and which exhibits emergent intelligent behaviour such as adaptation, learning and co-evolution with their environment. The intelligence of the scheduler emerges from the horizontal and vertical interaction of its constituent agents balancing their individual and group interests.