PUB - Publikationen an der Universität Bielefeld

Robotic systems research is typically a result of a collaborative engineering process in an environment of rapidly changing technologies. It involves a large number of hardware and software components, each of which are problem solutions for differ- ent challenges from different research areas. Systems for Human-Robot Interaction face the additional challenge of having to actually work together with humans in a shared environment. Despite the fact that some single capabilities needed to solve various tasks with robotic systems are well established, however, robots frequently fail when they need to combine these capabilities and demonstrate them in a complex real world scenario. The aspect of coordinating and efficiently combining robots capabilities is one area where robots still fail. The focus of this work is to provide a framework for developers of interactive robot systems that perform in domestic environments, which allows the combina- tion and improvement of building blocks of the robot behavior based on experience gained in real world interaction and make the combination and coordination of these building-blocks easier and more easily reusable for developers. With the framework developed during this thesis there where many occasions where a robot platform was evaluated in a real world environment. This iterative design process documented here helps to answer questions about how to improve the robot performance based on observations from real world interactions, how to enable re-usability of robot behavior building blocks across scenarios and plat- forms and how to combine and coordinate the different robot capabilities from a developers point of view.