In recent years there is an increasing interest in building personal robots that are capable of a human-like interaction. In addition to multi-modal interaction skills, such a robot must also be able to adapt itself to unknown environments and, therefore, it has to be capable of knowledge acquisition in a lifelong learning process. Moreover, as humans are around, reactive control of the robot’s hardware is important, too. Consequently, researchers aiming to realize a personal robot have to integrate a variety of features. Due to the very different nature of the necessary capabilities, interactive robotics research is thus a truly interdisciplinary challenge.