Robots Who Interrupt Talk in Meetings

Knowledge sharing is an important aspect in most meetings. Personal characteristics of some participants, such as their (in)ability or (un)willingness to take the floor, may have a negative effect on the quality of knowledge sharing; some people tend to talk too much, while others have difficulties in making themselves heard. A robotic facilitator can be used to distribute the floor time more efficiently. While current research is mostly focused on encouraging participants to talk, this paper suggests interruption functionality to discourage speakers from talking. The facilitator gathers turn-taking signals from the participants and expresses them on their behalf. It hides the identity of individuals, making it easier for everyone to take action. The facilitator represents the signals coherently for all signalers, which equalizes the differences in social signalling skills, and makes it easier for the speaker to interpret the signals. It continuously gathers feedback from all participants, and thereby can represent the collective mood of the audience and smooth out outlier reactions. The facilitator can be programmed to act in a germane, courteous and attentive manner, which helps keeping the meeting mood high.