Controlled Emergence and Self-Organization

Summary

Admiration of nature’s ability to develop robust self-organizing and complex structures showing emergent behavior was the starting point for the Organic Computing (OC) endeavor. Although the concepts of self-organization and emergence have been subject to extensive investigations and discussions for more than 100 years, soon it became clear that we lack a quantitative assessment of these concepts as a basis for an implementation in technical systems. The main questions to be answered in this context are: Can we define emergence and self-organization (or sub-concepts thereof) compatible with a quantitative, experimental, and objectifiable method as required in natural science? Can we control self-organization and emergence without forcing their meaning? Are there generic architectures generally applicable to technical systems serving this purpose? In this chapter, we will try to give some answers to these questions. After an introduction and specificationv of the problem we will review some recent approaches to a definition of emergence and assess them with respect to their usability in our technical context. We will then introduce an architectural template, the Observer/Controller architecture, which seems to be a key feature in most OC systems. In addition to the general pattern –; essentially constituting a higher-level control loop –; this Observer/Controller architecture will be developed in some detail as a framework for own implementations. We present a quantitative approach for a technically relevant definition of emergence and self-organization, and propose a systematic approach to a ranking of various Observer/Controller architectures.