Framing Human-Robot Proxemics

Title:Framing Human-Robot Proxemics
Institute:University of Twente (HMI)
Place:Enschede The Netherlands
Type:final project
Start date:1 januari 2017
End date:not present
HMI ContactJered Vroon


Can we use framing/priming to influence what is perceived as an ‘appropriate’ distance for a robot?

Much of the social robotics research at the University of Twente focuses on mobile robots having conversations with people. Even though the applications are diverse (e.g. airport robot SPENCER, museum tour guide FROG, telepresence robot TERESA), a central question is that of social positioning; How should a social mobile robot position itself with respect to the people with which it interacts? Which distance should a robot keep to the people with which it tries to have a conversation?

Proxemics, first coined by the sociologist Edward Hall, provides a deceptively direct answer to this question. Within it, a range of social distances has been defined for different levels of intimacy. For example, the distance for interaction with acquaintances would be between 1.2 and 3.6 meters.
For this reason, proxemics has been actively used in a variety of mobile social robots. Some authors simply use a distance of approximately 1.2 meters as a minimum for their approach and navigation behaviour. Others have looked at different factors that influence the appropriate proxemic distance, such as ‘familiarity with the robot’, ‘gender’, ‘properties of the robot’, and ‘history of pet ownership’.

At the same time, these findings, along with a variety of findings from psychology, suggest that there is more to proxemics than meets the eye; internal factors such as familiarity and self-construal can apparently also have an effect on what is appropriate proxemic behaviour. If that is the case, it should be possible to change what is perceived as appropriate proxemic behaviour of a robot with the right priming/framing. And if that is indeed the case, an interesting argument can be made against straightforwardly applying static proxemic distances to robotics.

Your assignment, should you choose to accept it, will be to design and conduct clear and clean experiments into the effect of framing/priming on human-robot proxemics. Your work will be embedded in a larger effort within the TERESA project.

The applicant is expected to have a strong background in methodology. Programming intelligent robot behaviour will not be necessary, since you can instead remote control it (‘Wizard of Oz’-approach).