ON HOW A SELF-ORGANIZING SYSTEM PRODUCES COLLECTIVE BEHAVIOR

In this paper we address the following question: How can the designers of self-organizing systems use the relationships among relevant parameters to quantitatively characterize how a self-organizing system produces the collective behavior process, to better control and design a self-organizing system? A self-organizing system has advantages in performing dangerous and exploratory tasks that are not suitable for humans. However, it is often difficult for designers to design self-organizing systems, because, in an environment with dynamic complexity, the process of how self-organizing systems behave collectively is unknown. To address this difficulty, we propose a method to quantitatively represent environmental complexity as well as the system behavior, and study the relationship between them so as to gain insights into how a self-organizing system produces collective behavior. In this paper, we identify four collective behavior patterns that a self-organizing system produces when the environment complexity changes, namely, initialing pattern, adjusting behavior, stabilizing pattern, and restarting pattern. We illustrate the efficacy of our method using a box-pushing problem. Our focus is on describing the method rather than the results.