Dynamic Integration of Ubiquitous Robotic Systems through Capability Model Processing

The promise made by ubiquitous robotics to provide robotic functions anywhere and anytime would raise two major issues: the integration and the management of a complex system that could include a potentially large amount of devices of different natures. While we define four levels of abstractions at which the integration issue should be addressed, we argue that a successful dynamic integration of ubiquitous robotic artifacts must rely on semantically rich models. We then describe how, from a fundamental semantic model, we derive semantic capability models of these artifacts that can in turn be used at the different stages of the artifact lifecycle, from the generation of the code that would realize the capability to its implication into a usage scenario. Finally, we describe the application of our approach to the development of a ubiquitous robotic home environment