What feels a comfortable and appropriate space between ourselves and another, is a nonverbal, and powerful element during social interaction. Research on social space, sometimes known as proxemics, suggests that the proximity between conversation partners may inform the nature or quality of a relationship between them. However, what feels an optimal distance seems to vary somewhat according to participants’ cultural backgrounds as well as other demographic profiles such as gender or height.
Proxemic experience during social interaction is thus likely to involve a robust yet complex set of social signals. To what extent our experience during human-human interaction may be shared during our interaction with robotic agents, is an important and emerging field of study in the human-robot interaction (HRI) literature. We present pilot data from the UK arm of our on-going UK-Japan cross-cultural project, ProxiBots study. Our initial plan for the behavioural study was converted into an online study to adapt to laboratory closures and travel restrictions due to COVID. Forty-four adults who were UK nationals and have primarily lived in the UK, watched 36 video clips of NAO robot walking towards them within a virtual setting. Each video clip displayed a marked point at which participants were asked to mentally place themselves in relation to Nao robot on approach. NAO robot walked either straight, from the right or left of this marked point in an indoor or outdoor setting. The walking speed of the robot was either at its manufacturer’s default speed or twice or four times faster via video editing. Visualisation and modelling analysis, using linear mixed effects modelling (lme, R), suggested that approach angle played a role in our decision on how close a robot should approach us. We will discuss the results as to what extent they may resonate with earlier studies during in-person HRI, as well as in the context of our plans for next steps in the project.