Haptic collaboration is defined as the mutual exchange of force and position signals between partners to achieve a shared goal. Haptic collaboration has broad applications in robotics, e.g. when robots represent virtual partners in dancing, when robots help handicapped people by guiding them, in service robotics and in multi-robot telepresence scenarios.

The presented research is based on the assumption that human-robot-interaction is more successful when robots can behave in a human-like way. Therefore human-human-collaboration (HHC) needs to be studied experimentally with the goal to model HHC and then substitute one partner by a technical system. For this experimental analysis, the complexity of the aforementioned scenarios has to be reduced: a virtual pursuit tracking task for two partners as a new paradigm was developed in this line of research. This experimental paradigm allows to measure differences between the intended and actually performed trajectory of the jointly manipulated object.

Haptic collaboration is studied within a framework separating between low and high-level interaction in dependence of the involved cognitive abilities. So far the following constructs were investigated on a lower level:

1. Efficiency of haptic feedback
2. Dominance difference between the two partners of a dyad
3. Consistency of behaviour across time and partners

Ongoing research deals with decision making which represents higher level haptic collaboration.