Approaching human-like spatial awareness in social robotics: an investigation of spatial interaction strategies with a receptionist robot

Abstract

This doctoral thesis investigates the influence of social signals in the spatial domain that aim to raise a robot’s awareness towards its human interlocutor.\nA concept of spatial awareness thereby extends the robot’s possibilities for expressing its knowledge about the situation as well as its own capabilities.\nAs a result, especially untrained users can build up more appropriate expectations about the current situation which supposedly leads to a minimization of misunderstandings and thereby an enhancement of user experience.\nOn the background of research that investigates communication among humans, relations are drawn in order to utilize gained insights for developing a robot that is capable of acting socially intelligent with regard to human-like treatment of spatial configurations and signals. In a study-driven approach, an integrated concept of spatial awareness is therefore proposed. An important aspect of that concept, which is founded in its spatial extent, lies in its aspiration to cover a holistic encounter between human and robot with the goal to improve user experience from the first sight until the end of reciprocal awareness. It describes how spatial configurations and signals can be perceived and interpreted in a social robot. Furthermore, it also presents signals and behavioral properties for such a robot that target at influencing said configurations and enhancing robot verbosity.\nIn order to approve the concept’s validity in realistic settings, an interactive scenario is presented in the form of a receptionist robot to which it is applied. In the context of this setup, a comprehensive user study is conducted that verifies the implementation of spatial awareness to be beneficial for an interaction with humans that are naive to the subject. Furthermore, the importance of addressing an entire encounter in human-robot interaction is confirmed as well as a strong interdependency of a robot’s social signals\namong each other.