The context of this paper is the design of an appropriate working environment for the development of learning robots. We are interested in robots which are able to structure incoming sensor information on the basis of internal reference schemes and robot tasks. For efficiency reasons, we want to perform the learning cycles in simulated environments explored by the robot. In this paper, we outline how advanced visualization and interaction techniques as developed in the field of Virtual Reality could be explored to study the development and properties of the internal data of a semi-autonomous robot, as well as the learning process itself.