Flexible planning and control of a factory requires concepts for the representation of the local and global knowledge of the autonomous systems working in teams. Therefore, the author developed a distributed real-time knowledge base that facilitates team work of autonomous systems in flexible manufacturing environments. The knowledge base consists of several local knowledge bases cooperating with each other. In each local knowledge base the knowledge is represented as an object structure with demons. If several autonomous systems cooperate to perform a certain task, the local knowledge bases of these autonomous systems form a team to provide the knowledge needed by the autonomous systems. First applications with single local knowledge bases have been successfully implemented.>

Flexible planning and control of a factory requires concepts for the representation of the local and global knowledge of the autonomous systems working in teams. Therefore, we developed a distributed real-time knowledge base that facilitates teamwork of autonomous systems in flexible manufacturing environments. The knowledge base consists of several local knowledge bases cooperating with each other. In each local knowledge base the knowledge is represented as an object structure with demons. The local knowledge bases provide multi-user and real-time access. Each local knowledge base has an active component which asynchronously passes changed data to interested application processes. If several autonomous systems cooperate to perform a certain task, the local knowledge bases of these autonomous systems form a team to provide the knowledge needed by the autonomous systems. The local knowledge bases cooperate by means of selective updating and location-transparent access to the global knowledge. First applications with single local knowledge bases in the fields of image interpretation, navigation and task planning have been successfully implemented.>

KONFGEN, a system that speeds up the work of building and updating expert systems, is discussed. It generates expert systems for configuration tasks (e.g. configuration of machines) from a formal, constraint-based specification of the machine's components and their interdependence. A formal language which is used to specify a configuration problem, the different kinds of constraints the language supports, the evaluation of those constraints by the expert system generated, and how KONFGEN generates an expert system from a problem specification are described.>