Learning coordination skills for service robots

Currently, an important topic in robotic research systems is the design and development of learning Multi-Agent Systems (MAS). One major advantage of these systems is the fact that several agent work towards a common goal, having different specializations for specific subtasks. Especially, complex systems as service robots consist of several autonomous modules capable of for example moving, seeing and collision avoidance. For successful task solution cooperation between these agents is required. Since effective cooperation between these different agents can be extremely difficult to program, one possibility to cope with this problem is to learn these cooperation strategies. In this paper a formal agent model with an effective coordination model will be presented. It will be shown how the system's cooperation strategy can be learned on human instruction and advice. Copyright (C) 1999 IFAC.