Due to reduced prices and system complexity, Motion Capture (MOCAP) technologies, which traditionally are applied in ergonomics research and film/game industry, nowadays find more and more use in other areas (Bergler 2007, Brodie et al., 2008, XSens 2012). Another new technology, which finds application in an extended set of areas,like online markets or computer games, is the provision of instructions by animated virtual characters in order to successfully support a multitude of different users. Yet, when considering maintenance tasks, there are only few examples (Ziegeler & Zühlke, 2005). A well-known one is the animated pedagogical agent “Steve”, who teaches users how to use or overhaul complex machines (Rickel & Johnson, 1999). According to Rickel (2001) many advantages are united in animated pedagogical agents such as attention guidance with gaze and gestures. The major benefit is a deeper and faster learning process especially of procedural knowledge. This results from synchronizing oral explanation with visualized action sequences and with that a much more facilitated transfer of practical knowledge (Mischas & Berry, 2000). Regarding the context of Industrial Product-Service Systems (IPS²), which are characterized by a variety of different users, contexts of usage and highly specialized products and services an urgent need of descriptive and comprehensible user support is clearly noticeable (Uhlmann et al., 2008, Schmuntzsch & Rötting, 2011). Thus, our IPS²-related project (SFB/ TR29 B4) focuses, inter alia, on the combination of MOCAP technology with the modeling of an animated pedagogical agent, who demonstrates how to perform a maintenance task in an instruction video. In this article we will give a brief overview of the design process and address the advantages and challenges which came up.