In this paper, we discuss some phenomena of object clustering by distributed autonomous robots. This work was motivated by Swiss Robots, which collect scattered objects into some clusters without any global information nor intelligent concentrated controller. The clustering mechanism will be generated from the interaction among the mechanical system (i.e. body), the control law (i.e. brain), and the environment (i.e. real world). In order to understand the clustering mechanism, we utilize a “Implicit Control Theory”. This theory proposes not only “Explicit Control Law”, which corresponds to general control law, but also “Implicit Control Law”, which reflects the interaction among the body, the brain, and the environment. Based on the “Implicit Control Theory”, the object clustering by Swiss Robot is investigated. Moreover, this paper investigates not only the Swiss Robot's object clsutering, but also object clsuterings by another simple robots (Coronoc Robot, and Aggregator Robot). Though the Explicit Control Law and mechanical structures of each robot are exactly different, we can see common parts (principles) of Implicit Control Law. Furthermore, we demonstrate the correspondance between the Explicit Control Law and its clustering ability of each robot (e.g. number of clusters, clustering position).
