An increased use of robots in direct interaction with humans can be observed both in the industry and in the service sector. The specific areas of application as well as forms of interaction are diverse. Nevertheless in order to enable a structured comparison, a taxonomy is necessary, that, on one hand is generic enough to be applicable to a wide variety of human-robot interaction scenarios and on the other hand is specific enough to enable the analysis of superordinate features. This article introduces the development of a taxonomy that meets both criteria. A taxonomy of human- robot interaction was developed, which both meets the requirement of the highest possible generalizability and enables the interaction characterization of humans and robots with a high level of detail. The influence of certain design features of the robot on typical human factors aspects such as trust, acceptance, situation awareness or mental stress can be subjected to a systematic (meta) analysis. The taxonomy thus forms the basis for future recommendations for innovative and human-centered activities.