Deciding when and how to touch a human: social technologies for human-machine interaction

Fabien BOUCAUD, Post-doctoral fellow at ISIR (Institute of Intelligent Systems and Robotics, UMR CNRS 7222), France

With the latest developments in the field of affective computing, we are seeing the appearance of virtual characters with increasingly advanced social interaction capabilities: speech, gestures, relational strategies,... However, despite its importance in our daily social lives and in our general well-being, social touch is a modality that is still not well integrated in new communication technologies. Setting up an interactive loop between a human and a system capable of making touch decisions is a challenge that requires asking questions of both a technical and social, philosophical or ethical nature. In particular, how to create a computational model of social touching behavior? How to make a human feel a nuanced and appropriate touch? Which touches are acceptable in which situations? How to touch a human in a credible and consistent way? A first implementation example and a first evaluation of an autonomous touching agent will shed light on human behavior and reactions to these new social technologies.