388
V. Vouloutsi et al.
The proposed experimental setup was conceptualised to resemble as much
as possible a real-world scenario, where the robot would be placed in a public
setting, navigate and look for people to take a selfie. A humanoid robot creates
natural curiosity to participants who immediately approached the robot, and
this prevented the robot from showing all the intended behaviours (like getting
the attention of a participant) correctly. Additionally, all interactions were much
closer than what we originally anticipated. Novelty and proxemics could have
affected the way participants perceived the robot’s behaviour.
Due to the small sample size, this study has just scratched the surface of
what could be the relevant factors in HRI for group social settings, such as
taking a selfie. Many improvements can be suggested for the interaction models
of the selfie behaviour. We are currently working on incorporating a sophisticated
cognitive architecture that can provide more robust interactions and improving
the exhibited gestures.
Perhaps the most important reflection of the conclusion is that while HHI is a
viable basis on which to base the design of HRI, HRI at this point in time is still
fundamentally different than HHI. Studies such as this one help to fill the gap in
research for some particular social contexts and can thus highlight what specific
factors may influence HRI regarding a specific task. Ultimately the underlying
basis and research question for this study, which sought to show that the same
theories of social cognition in HHI can be applied to HRI, has been shown to
be true. The robot’s ability to successfully take a group selfie in the majority of
cases displaying only social sensorimotor information, suggests that socSMCs are
a powerful vehicle for conveying intention of a robotic agent in an interaction.
However, the realities of the study have also shown that simply considering a
social context in HHI and applying it directly to HRI cannot always be done.
This is evident when considering the HHI selfie-taking observational pre-study,
in which humans were shown to be incredibly shy and sceptical of strangers
approaching them for any reason (especially for a selfie) contrasting with the
final experiment where the humans exhibited no signs of scepticism towards the
robot. This study contributes to the field by exploring the use of socSMCs and
their link to personality in a social group setting and how they can be utilised
to facilitate transparent interactions.
Acknowledgments. This work is supported by grants from the European Research
Council under the European Union’s Socialising Sensori-Motor Contingencies (socSMCs) grant agreement n. 641321.
References
1. Dautenhahn, K.: Socially intelligent robots: dimensions of human-robot interaction. Philos. Trans. R. Soc. B: Biol. Sci. 362(1480), 679–704 (2007)
2. Knapp, M., Hall, J., Horgan, T.: Nonverbal Communication in Human Interaction.
Cengage Learning, Boston (2013)
3. O’Regan, J.K., No¨ e, A.: A sensorimotor account of vision and visual consciousness.
Behav. Brain Sci. 24(5), 939–973 (2001)
V. Vouloutsi et al.
The proposed experimental setup was conceptualised to resemble as much
as possible a real-world scenario, where the robot would be placed in a public
setting, navigate and look for people to take a selfie. A humanoid robot creates
natural curiosity to participants who immediately approached the robot, and
this prevented the robot from showing all the intended behaviours (like getting
the attention of a participant) correctly. Additionally, all interactions were much
closer than what we originally anticipated. Novelty and proxemics could have
affected the way participants perceived the robot’s behaviour.
Due to the small sample size, this study has just scratched the surface of
what could be the relevant factors in HRI for group social settings, such as
taking a selfie. Many improvements can be suggested for the interaction models
of the selfie behaviour. We are currently working on incorporating a sophisticated
cognitive architecture that can provide more robust interactions and improving
the exhibited gestures.
Perhaps the most important reflection of the conclusion is that while HHI is a
viable basis on which to base the design of HRI, HRI at this point in time is still
fundamentally different than HHI. Studies such as this one help to fill the gap in
research for some particular social contexts and can thus highlight what specific
factors may influence HRI regarding a specific task. Ultimately the underlying
basis and research question for this study, which sought to show that the same
theories of social cognition in HHI can be applied to HRI, has been shown to
be true. The robot’s ability to successfully take a group selfie in the majority of
cases displaying only social sensorimotor information, suggests that socSMCs are
a powerful vehicle for conveying intention of a robotic agent in an interaction.
However, the realities of the study have also shown that simply considering a
social context in HHI and applying it directly to HRI cannot always be done.
This is evident when considering the HHI selfie-taking observational pre-study,
in which humans were shown to be incredibly shy and sceptical of strangers
approaching them for any reason (especially for a selfie) contrasting with the
final experiment where the humans exhibited no signs of scepticism towards the
robot. This study contributes to the field by exploring the use of socSMCs and
their link to personality in a social group setting and how they can be utilised
to facilitate transparent interactions.
Acknowledgments. This work is supported by grants from the European Research
Council under the European Union’s Socialising Sensori-Motor Contingencies (socSMCs) grant agreement n. 641321.
References
1. Dautenhahn, K.: Socially intelligent robots: dimensions of human-robot interaction. Philos. Trans. R. Soc. B: Biol. Sci. 362(1480), 679–704 (2007)
2. Knapp, M., Hall, J., Horgan, T.: Nonverbal Communication in Human Interaction.
Cengage Learning, Boston (2013)
3. O’Regan, J.K., No¨ e, A.: A sensorimotor account of vision and visual consciousness.
Behav. Brain Sci. 24(5), 939–973 (2001)