AAAI 2016 SPRING SYMPOSIA

Enabling Computing Research in Socially Intelligent Human-Robot Interaction: A Community-Driven Modular Research Platform

Advances in sensor and communication technologies have facilitated progress in computing research on physical platforms. The field of human-robot interaction (HRI) has grown significantly in the last decade and a half, and actively brings together an interdisciplinary community of researchers across computing, AI, robotics, and social science. However, progress has been limited by the lack of affordable, general-purpose, modular hardware robot platforms with available low-level software that would enable large numbers of computing researchers to enter the field and develop and test algorithms, as well as conduct statistically significant user studies by deploying systems in the real world and collecting user data to inform further computational research in HRI.

The goal of this symposium is to kick-off the process of community-informed design and development of the low-cost HRI hardware and software platform, to be developed by the symposium organizers, with NSF support. The hardware design will involve advances in user-centric yet affordable design, safety, modularity, generality, and system integration. The software design will involve novel general algorithms and open source robust code bases that enable the hardware platform to operate “out of the box” with a set of socially intelligent behavior primitives enabling computing researchers to focus on their areas of interest without having to develop low-level robot control algorithms and code.  Both the robot hardware and software necessary for pursuing the computational, AI and non-contact HRI challenges have unique requirements driven by the need to be socially aware and socially expressive. The resulting platforms must be capable of recognizing social signals, reasoning over those signals, and generating appropriate behaviors in response.

The symposium will present initial hardware design ideas and plans, along with exploratory exercises to determine the usability of proposed software systems as well as the fit of capabilities with the community’s needs. Our “design by quorum” is combined with modular design that centers on creating a standard vetted by the community and builds on recent technologies to minimize cost. The symposium will address computing challenges that bridge AI, human-computer interaction (HCI), service robotics, and other related areas. Therefore, advances made through the discussion at the symposium will serve to push the field forward, thereby impacting the computing community at large, including AI and robotics.

Topics

Symposium topics include, but are not limited to the following:

  • Recognition and generation of fundamental social behaviors, such as spacing (i.e., where to be), eye gaze (i.e., where to look), natural language (i.e., what to say), body language (i.e., how to act), and timing (i.e., when to act), among others

  • Dialog/interaction management, decision-making, and learning

  • Computational models of social dynamics and interaction patterns in human-robot interactions

  • Mapping, localization, path-planning, and navigation in human environments

  • Context/situational awareness and scene understanding in human-robot social interactions

  • Online adaptation to human social behavior and interaction contexts

  • Long-term learning of human behaviors, preferences, and needs

  • Software architectures, tools, and systems for facilitating human-robot interactions

  • Sensor, mechanical, and computational hardware for enabling human-robot interactions

  • Ethics in the design of social robot hardware and software

Format

The symposium will be a combination of presentations, posters, invited talks, plenary sessions, and breakouts, to maximize participant interaction:

  • Paper presentation sessions for accepted full papers on the topics of (1) social robot design and applications, and (2) computational methods and software for robust social robot behaviors

  • Poster sessions for accepted short papers and position papers on topics similar to those in the paper presentations

  • Demos (optional) for accepted full papers, short, and position papers

  • Plenary talks by invited speakers

  • Panel discussions on the topics of (1) the design of social robots to support strong computational methods and applications, and (2) the design of robot behaviors and software to further HRI research

  • Breakout sessions in which participants will (1) design an ideal social robot platform to support strong computational methods and applications; (2) formalize robot behaviors and software architectures for robust HRI systems; (3) prototype social robot hardware, software, and/or applications given hands-on materials; and (4) discuss new collaborative research efforts

Submissions

Prospective authors are invited to submit full papers (6-8 pages) and/or short/position papers (2-4 pages) in PDF format to EasyChair. Accepted papers will be published in a technical report on the AAAI Digital Library

Organizing Committee

Dr. Maja Matarić, USC, mataric(at)usc.edu

Dr. Mark Yim, UPenn, yim(at)grasp.upenn.edu

Ross Mead, USC, rossmead(at)usc.edu

Important Dates

Conference:

March 21–23, 2016

Symposium:

TBA

Associated Links

AAAI Conference Page

Modlab at UPenn

Interaction Lab at USC