Making Appearances: How Robots Should Approach People

Michiel Joosse, Manja Lohse, Niels van Berkel, Aziez Sardar, Vanessa Evers

Research output: Contribution to journalJournal articleResearchpeer-review

19 Citations (Scopus)
69 Downloads (Pure)

Abstract

To prepare for a future in which robots are more commonplace, it is important to know what robot behaviors people find socially normative. Previous work suggests that for robots to be accepted by people, the robot should adhere to the prevalent social norms, such as those related to approaching people. However, we do not expect that socially normative approach behaviors for robots can be translated on a one-on-one basis from people to robots, because currently robots have unique and different features to humans, including (but not limited to) wheels, sounds, and shapes. The two studies presented in this article go beyond the state-of-the-art and focus on socially normative approach behaviors for robots. In the first study, we compared people's responses to violations of personal space done by robots compared to people. In the second study, we explored what features (sound, size, speed) of a robot approaching people have an effect on acceptance. Findings indicate that people are more lenient toward violations of a social norm by a robot as compared to a person. Also, we found that robots can use their unique features to mitigate the negative effects of norm violations by communicating intent.

Original languageEnglish
Article number3385121
JournalACM Transactions on Human-Robot Interaction
Volume10
Issue number1
Number of pages24
ISSN2163-0364
DOIs
Publication statusPublished - 2021

Keywords

  • Social robots
  • functional noise
  • height
  • human-robot collaboration
  • human-robot interaction
  • interpersonal distance
  • personal space invasion
  • proxemics
  • robot approach
  • robot navigation
  • social norms
  • velocity

Fingerprint

Dive into the research topics of 'Making Appearances: How Robots Should Approach People'. Together they form a unique fingerprint.

Cite this