Possibilities and limitations of the application of agent systems in control

Distributed control systems are an overwhelming trend in industrial control on all level of control. One possible technology for implementing such distributed control systems are agent systems and the agent programming paradigm. This technology will open up a free of breakages architecture. But the applicability of these systems within the different levels of control is different. Within this paper the general agent system architecture and the requirements of different control levels are characterised. Based on this possibilities and limitations of agent application within control is sketched. The paper will open up the view of practitioners and end-users on the agent system application beyond enterprise planning and manufacturing execution as a mean to implement a free of breakages factory spanning control architecture.