Reactive Self-Configuration of an Ecology of Robots

The field of ubiquitous robotics is burgeoning, and different brands of massively distributed heterogeneous robotic systems are being proposed and applied to several domains. The strong added value of these systems comes from their potential ability to dynamically self-configure, by changing the form of their cooperation to adapt to a given task or situation. In face of this, no satisfactory solution exists to the problem of how such a system should self-configure. In this paper, we explore a reactive approach to self-configuration inspired by ideas from the field of semantic web services. We illustrate our approach on a specific type of ubiquitous robot system, called a PEIS Ecology. We show experiments in which our approach autonomously generates a configuration to perform a cooperative navigation task, and dynamically changes this configuration when one of the components fails.