Personal tools
You are here: Home ACII ACII2009 Social interaction with robots and agents: Where do we stand, where to do we go?

Social interaction with robots and agents: Where do we stand, where to do we go?

ACII 2009 Special Session organized by Eva Hudlicka, Sabine Payr, and Rodrigo Ventura (chair: Sabine Payr)

Robots and agents are becoming increasingly prominent in everyday life, e.g. as companions, user interfaces to smart homes, household robots, or for lifestyle reassurance.

In these roles, they have to interact with their users in a complex social world, and should build and maintain long-term relationships with them. The European FP7 project "Social Engagement with Robots and Agents" (SERA, http://project-sera.eu/) has set out to make progress in the study of social interaction with such artefacts, taking a fresh and application-oriented look at the requirements and models, but also at the methods with which to develop theories and models.

Much research has been done already on prerequisites for, and modelling of, social robots and agents. The complexity of the task sometimes seems overwhelming, however. The aim of this special session is to address some fundamental questions:

  • What do we know already about social interaction with agents and robots, and about the role of emotions in it?
  • How can existing theories and data be of use in advancing our knowledge? How should we study them to extend our knowledge of social interaction with robots and agents?
  • Which additional data, experiments, disciplines and methods should be called upon?
  • Are existing approaches to affective modelling capable of modelling both episodic and pervasive emotions in social agents and robots? What additional requirements and methods may be required to model these emotions?
  • Do we know enough to model systems that build and maintain long-term relationships with users? Are existing theories sufficient to explain what happens in this kind of social interaction, and existing models sufficient to generate it? What types of additional theories and models would be required? What types of data would be required and what are the best means of obtaining such data?

Format:

  • Welcome, outline of the session's goals, introduction of panellists, and projection of video sequences.

    The project SERA sets up a field study in which subjects interact with a robotic user interface (the Nabaztag - http://www.nabaztag.com), enhanced with some sensor and ASR technology in their homes, over the period of one week. Video data from the first round of the study with the "baseline system" will be available in time for ACII. We will show an extract and contrast it with a video of human-human interaction. The videos will be made available to the panellists well before the conference so that they can watch, analyse, comment, and refer to them in their presentations.

  • Statement of challenges, questions to the panellists

    The panelists will have been given those questions in advance, so that they can discuss and refer to them in their presentations.

  • Brief presentations by the 4 invited speakers, addressing the questions and challenges, and referring to the video sequences
  • Discussion among panelists only
  • Questions and discussion contributions from the audience

Panelists

Christian von Scheve, Christian Becker-Asano, Ana Paiva (Kerstin Fischer - excused)
Document Actions

Welcome to the discussion forum!

Posted by Paolo Petta at Monday, 12 October 2009
This page serves to collect enquiries and comments, and generally to continue the discussion concerning the topic of the Special Session presented at ACII 2009.
We would like to thank the chairs for their kind availability!

Paolo, Nadia and Cai

"It reminds me of my grandmother!"

Posted by Sabine Payr at Thursday, 15 October 2009
Another participant made an extremely interesting comment on the Nabaztag video. While most participants couldn't help laughing (mostly out of irritation, I suppose) at the situation and the dialog, he said that it reminded him of the interactions he experienced as a child, at home: his grandmother used to interrogate him in the morning in this style. Although the interaction in itself was no more friendly or "successful" as the one in the video, the memory was precious to him: it was embedded in the larger context of "feeling at home" and being nurtured and cared for. <p>
This statement brought into the debate an aspect of long-term relationships which has hardly been explored yet in affective computing: it is the aspect of habit-building and the establishment of rituals that by themselves signify the relationship. One may not or need not like them, but one misses them once the relationship is broken.

"Who is companion to whom?"

Posted by Sabine Payr at Thursday, 15 October 2009
One participant pointed out that in the Nabaztag video, the roles seem to be inverted: the human answering the questions appears as the one who "serves the needs" of the companion, not the other way round. <p>
That led on to a discussion in which "serving the needs" of the other was pointed out as an important building block of long-term relationships, and can be seen as a special case of "mutual investment". Example: one likes one's dog also because it has to be fed, walked etc., and because one has invested time and energy into educating it.


Powered by Plone

Portal usage statistics