Among the many ways of communication between humans, touch is one of the most powerful. Touch is often used by instructors in sport or dance classes to adjust the students posture in a very intuitive way. Using tactile instructions appears thus a very appealing modality for developing humanoid robot motions as well. Interpretation of tactile instructions spontaneously given by users is a complex task for artificial systems. Furthermore, the way users employ touch to communicate their intention is still completely unexplored. This work reports the first results obtained in the realization of a...