| SOCS: Self-Organizing Control Systems |
|
Organization responsible: TNO People involved:
Project description: This project explores the use of biological self-organisation principles in networked systems for increased levels of robustness, flexibility and adaptivity in control systems. Most man-made technological systems rely on an intelligent operator. The whole system is controlled by this operator in a centralised manner. This centralised control approach has been the model for the design of most current automated control systems. However, centralisation comes at a price: it introduces single points of failure, hence decreasing robustness. In sharp contrast, many natural (biological) systems are regulated by their own internal processes, lack centralised control and yet display very complex behaviour. Many natural systems system are consequently known to be self-organising, and characterised by their autonomy, robustness and adaptivity. The way that behavioural and structural patterns emerge from the interaction of many simple elements, is a complex phenomenon that intrigues scientists from all disciplines. Not surprisingly, the principles behind these natural systems have made the transition to systems engineering as well. Biologically inspired systems are becoming more and more widespread in society, for instance in logistic applications, surveillance systems and data communication networks. This project focuses on applying biological self-organisation principles in agent-based systems in order to achieve new levels of robustness, flexibility and adaptivity for control systems. The focus of the SOCS project is to research self-organizing capabilities in networked systems. The circumstances under which a self-organizing system operates may differ for various application areas. Therefore, we want to investigate how a number of crucial aspects influence the effectiveness and the efficiency of the team-performance of self-organizing systems. The result, an understanding of how these aspects influence the system behaviour, will yield us generic knowledge required for engineering Self-Organizing Systems for different challenges. Publications: |