Raamwerk voor Reactieve Dynamische Resource Scheduling: Intelligente Capaciteitsbenutting in een Gedistribueerde Oving (Rabbit)
This project studies the development of systems for reactive scheduling of resources with adaptive behavior in a dynamically evolving environment. A central aspect of this project is therefore intelligent use of limited resource capacity. The specific limitations of small, mobile and autonomous devices with variable degrees of connectivity ultimately determine the possibilities of such systems. The definition and recognition of this context, and the ability of autonomous agents to reconfigure themselves in order to take advantage of this, should lead to a coherent application behavior and effective usage of the available resources.
Since this project involves dynamic systems that are mostly reactive, it is usually not possible to reach optimal conditions within the timeframe available for a response. In addition, the information available is typically incomplete. This situation has led us to an agent-based approach in which each agent considers a part of the global system from its own perspective and tries to accomplish its own goals after interacting with other agents. This topic is studied using a number of representative case studies.