Controlling Software-Induced Self-Organizing Behavior

Self-organization is an elegant mechanism to allow systems to adapt to a changing execution environment and to the various challenges that occur during the system’s lifetime. However, self-organization induces skepticism on the side of the users and even the developers of such systems, especially in safety-critical domains, since the system acts autonomously and is hard to predict. Therefore we present in this article an approach to develop self-organizing systems and to control this self-organizing behavior. The approach contains software engineering methods to construct such systems as well as formal methods to guarantee functional correctness and to measure the benefit of self-organization compared with traditional systems.
published 07.04.2015 in: Berlin, Germany Evolutionary Robotics, Organic Computing and Adaptive Ambience: Epistemological and ethical implications of technomorphic descriptions of technologies / [ed] Michael Decker, Mathias Gutmann, Julia Knifka, Year 2015

Publisher: LIT Verlag

ISBN: 978-3643901514