Engineering Organic Computing Applications—Design, Analysis, Safety

The term Organic Computing here stands for a number of techniques to incorporate nature inspired properties into technical systems (nothing to do with organic material in context of computing). One interesting class of Organic Computing (OC) applications are software intensive (embedded) systems with self-x properties. They are able to dynamically adapt to changes in requirements, to automatically detect and neutralize component failures and to continuously optimize themselves for better performance. These capabilities are called self-adaptation, self-healing and self-optimization. The expectation is that systems with self-x capabilities are by far superior to standard systems in terms of availability, reliability and effort of maintenance. Self-x embedded systems might have an enormous impact on innovations in production automation, automotive systems, space, robotics, sensor nets and many other fields. On the other hand self-x (or organic) embedded systems require new methods with respect to construction paradigms, functional correctness, safety and trust. One major challenge is to reconcile the flexibility of self-x systems with the demand for guaranteed positive (or at least non negative) in technical applications. Another challenge is to measure the degree of self-x capacity available in a technical solution, and finally a software engineering process for OC systems is needed. This part of the tutorial presents an approach to solve these problems. We introduce a canonical system architecture separating the productive system from the self-x infrastructure, and the Observer/Controller part of an OC application. Furthermore, we show how to achieve safety by model based analysis, and presented a technique how to compute the self-x capacity embodied in an OC application. It can be shown, that self-x mechanisms can indeed add new potential to classical redundancy- and fault-tolerance techniques. In addition, quantitative safety analysis helps to classify systems according to standard safety criteria (e.g. SIL).

Return to index

Last updated by tbourke at Mon Nov 26 15:38:09 2007 GMT+1100