Natural interaction based on affective robotics for multi-robot systems

In this paper we introduce a novel approach that enables users to interact with a multi-robot system in a natural manner. Additionally, interaction adapts to the user's affect and stamina. To this end, we consider the use of a smartwatch for recognizing the motion of the user's forearm, which is then translated in velocity commands for the robots. A natural mapping between user's movements and robots commands is implemented, so that the user can intuitively drive the robots by mimicking real-world behavior. Additionally, the operator's heart rate is measured, since it is a measure of mental stress. Thus, when an increase in mental stress is detected, the behavior of the multi-robot system is changed in order to simplify the interaction task and relieve the user. The proposed interaction system is tested with different users. Specifically, the effectiveness of the natural interaction is evaluated and compared to the use of a joypad and a haptic device. Moreover, we preliminarily test the affective interaction approach.