ETHNOS-a real time programming environment for robotic systems

Abstract

A great number of applications for robotic systems and, in particular, for mobile vehicles, require the ability to handle highly dynamic environments such as those typically involving human activity. In order to operate correctly, the systems are subject to many different timing constraints coming from many different sources: hardware, software and human intervention. In embedded systems, these real-time constraints are often dealt with ad hoc solutions that, as a consequence, cannot evolve or be adapted in any way. In contrast to this approach, the author believes that support both at the architectural, programming and operating system level, especially designed for this specific field of applications, is essential to the development of adaptive intelligent robotic systems. The paper tackles this problem, presenting a programming environment called ETHNOS (Expert Tribe in a Hybrid Network Operating System) that has been successfully used to develop autonomous robots operating in civil environments.