Mind the Gap: Proxemics and Personal Space Around Robots

Every person carries an invisible bubble of space, and how close a robot comes, where it looks and how much we like it all determine whether that bubble feels respected or invaded.

Hall's invisible zones

In the early 1960s the anthropologist Edward T. Hall coined the term proxemics for the study of how people use space in social life. He proposed that we carry a set of roughly concentric zones around us. Intimate distance, from touching to about half a metre, is reserved for close relationships. Personal distance, out to a little over a metre, is where conversations with friends happen. Social distance, up to three or four metres, suits colleagues and strangers, and public distance is the space between a speaker and an audience. Hall stressed that the exact boundaries vary by culture, and later work has added that they vary by individual, setting and mood.

Most of the time we manage these distances without noticing. We notice only when someone gets them wrong, by standing too close in a queue or hovering behind us at a desk. For robots that move among people, getting them right is not a nicety. A robot that blunders into personal space feels rude or threatening; one that keeps too far away feels aloof or broken.

Do people treat robots as social at all?

One of the first systematic attempts to apply Hall's framework to robots came from Michael Walters, Kerstin Dautenhahn and colleagues at the University of Hertfordshire in 2005. Participants approached a mobile robot and were approached by it, and the researchers recorded where they chose to stop. About 60 per cent of participants kept distances consistent with the zones people use with one another, which suggests they were treating the robot as a social entity. But a sizeable minority, around 40 per cent, came closer than the conventional personal zone, as if the robot were an object rather than a someone.

The team also looked at personality. Participants who scored higher on a factor the researchers labelled proactiveness tended to keep further from the robot. The effect was modest, but it pointed to an important idea: a single comfortable distance for all users is unlikely to exist.

Gaze, gender and experience

Leila Takayama and Caroline Pantofaru extended this picture in 2009 with a study testing people approaching a robot and being approached by one. They varied whether the robot's head was oriented towards the participant's face or towards their legs, and compared people with and without prior experience of robots. According to the summary in a later paper, they found that gender interacted with gaze: women were comfortable with a larger distance when the robot looked at their faces than when it looked at their legs, while men showed the reverse pattern. People who had experience with robots, and those who owned pets, were comfortable at closer distances.

The pet finding is intriguing. It hints that familiarity with non-human creatures that share our space, and sometimes invade it, makes people more relaxed about a machine doing the same.

Physical and psychological distance

Jonathan Mumm and Bilge Mutlu pushed the question beyond metres in a 2011 HRI study with the Wakamaru humanoid and 60 participants. They manipulated two things: whether the robot was likeable or unlikeable, and whether it maintained mutual gaze or looked away. Overall, people kept a greater distance when the robot held their gaze, consistent with a classic idea from nonverbal communication research that people compensate for too much intimacy in one channel by reducing it in another. But likeability changed the pattern. Participants who disliked the robot stepped back when it stared at them, whereas those who liked it did not change their distance.

The study also measured psychological distance, through how much personal information participants were willing to share with the robot. Those who disliked it disclosed less. The authors drew a practical conclusion: robots should establish some rapport before seeking closeness, and in public settings where rapport is unlikely they should be cautious with cues, such as sustained eye contact, that ratchet up intimacy.

Space as conversation

Taken together, this work suggests that proxemics is not a fixed buffer to be coded into a navigation system but part of an ongoing negotiation. Distance interacts with gaze, liking, personality, prior experience and gender, and each of these shifts what counts as polite. Engineers have built increasingly sophisticated socially aware navigation, often using models of personal space inspired by Hall, but the psychological research is a reminder that the same metre can feel friendly from one robot and invasive from another.

For social robots in homes, shops and hospitals, that means space should be treated as a signal in its own right. Approaching, pausing, angling the body or glancing away are all moves in a conversation that begins before anyone says a word.

References

1. Hall, E. T. (1963). A system for the notation of proxemic behavior. American Anthropologist, 65(5), 1003-1026. Link
2. Walters, M. L., Dautenhahn, K., te Boekhorst, R., Koay, K. L., Kaouri, C., Woods, S., Nehaniv, C., Lee, D. and Werry, I. (2005). The influence of subjects' personality traits on personal spatial zones in a human-robot interaction experiment. Proceedings of IEEE RO-MAN 2005, 347-352. Link
3. Takayama, L. and Pantofaru, C. (2009). Influences on proxemic behaviors in human-robot interaction. Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 5495-5502. Link
4. Mumm, J. and Mutlu, B. (2011). Human-robot proxemics: Physical and psychological distancing in human-robot interaction. Proceedings of the 6th ACM/IEEE International Conference on Human-Robot Interaction (HRI), 331-338. Link

Feature drafted with AI from the published research listed above, for social-robotics.com.