Personal Space Norms Aware Robotic Navigation Model and Its Evaluation in a Virtual Reality Environment

Abstract

In this study, we propose a robotic model that determines a robot’s ideal positioning when confronted with changes in personal space within a dynamic human-robot group. Recently, there have been several efforts to develop communication robots suitable for human communities. Determining a robot’s position is essential to avoid collisions with humans and allow the robot to maintain a socially acceptable distance from people. However, the interpersonal space maintained by people in a community may depend on the contexts and situation of the community. Thus, in a human–robot group, robots need to dynamically evaluate the changes made in their personal space and subsequently update their location while considering the positions of other group members. Here, we propose a robotic navigation model and evaluate its effectiveness in a virtual reality setting where a robot embedding the proposed model had to keep an appropriate distance from the other moving robots based on an experimental scenario. We examined whether experimental participants could distinguish the model-enabled robot’s trajectory from the other robot’s trajectory. The results showed that the robot embedding the proposed model could move in a humanlike way (convey an impression of group membership) by constantly finding a suitable position in the group, even when changes of personal space occurred. This suggests the meaningfulness of the proposed model toward enabling robots to behave as group members by selecting their pathway without violating the distance norms maintained by other human group members.