Day 3 Invited Talk | Paper List | Schedule | Social Events | Companion's Activities

Networked Fieldbus Systems
Dietmar Dietrich, Vienna University of Technology, Austria

Abstract

The roots of fieldbus technology are mixed, both classical electrical engineering and computer science contributed their share to the evolution. Typical performance of different fieldbus technologies overlap more and more. Nevertheless striking characteristic features are taking shape.

Although on one hand fieldbus technology has meanwhile established itself, on the other hand a large number of problems are still unsettled: the plain ISO/OSI model is no longer sufficient. We have to create new layers to obtain better interoperability. Furthermore, the dramatically increasing complexity of fieldbus applications demands scalable systems. The complexity together with high installation costs make simulations of fieldbus behavior a necessity.

Essential driving forces for the development of new concepts come from the area of building automation: the need to reduce the costs of integrating and automating facility management systems, service costs, and the need to reduce energy consumption while at the same time improving its distribution (demand side management systems). The prioritization of real-time data transmission combined with Gigabit switching standards for LANs offers new possibilities.

In order to achieve the goals of modern automation, new requirements have to be met: configurable interactions between different fieldbus systems, efficient wireless technologies to increase mobility and facilitate installation, but also plug & play features. A topic of increasing importance are the different possibilities for connections to IP-based LANs, which will partially replace conventional mid-level fieldbus systems. New concepts of implementation, tests and maintenance will make many of today's concepts obsolete. The network management, which has almost been neglected in the early days of fieldbus systems, is expanding at an incredible rate. Finally, we also have to think about agents in fieldbus systems and other concepts like HOLONic systems.

The comparison with biological systems forces us to adopt a new point of view. Will we be able to achieve the same redundant concepts? Of course, micromechanics, computer technology, and chemistry will create an endless number of new inexpensive sensors. Therefore, the number of fieldbus nodes in all systems will increase dramatically, similar to the distributed neural networks in living beings. No doubt, we are just at the dawning of a strong evolutionary development of smart systems.

Preliminary program subject to change without previous notice