Multidisciplinary Collaboration for Socially Assistive Robotics
From: AI Magazine - Fall 2008 - Page 98
By: Adriana Tapus

The 2007 AAAI Spring Symposium on Multidisciplinary Collaboration for
Socially Assistive Robotics brought together researchers in computer science,
engineering, sociology, and psychology to discuss this developing research
area. Socially assistive robotics focuses on the social, rather than the
physical, aspects of interaction between robots and humans in assistive
contexts (such as nursing homes, hospitals, rehabilitation, and home
environments). Human-robot interaction (HRI) for socially assistive
applications is a growing and interdisciplinary field that draws from a range
of disciplines in engineering, health sciences, social and cognitive
sciences, design, and the arts. Collaboration in this domain requires close
coordination and communication between diverse communities of practitioners
at all stages of the process: inception, design, development, use, and
evaluation. Challenges arise due to disciplinary differences in terminology,
methodology, practices, and ethical considerations inherent in
multidisciplinary collaboration. This symposium was a first attempt to bring
these multiple and diverse research communities together in order to promote
interdisciplinary discussion and knowledge transfer. Although this is a young
research area, systems are currently being developed for use in hospitals,
schools, and homes in therapeutic programs that monitor, encourage, and
assist their users. This is an important time in the development of the
field, when the broad technical community and the beneficiary populations
must work together to guide the intended impact of new technologies to
improve human quality of life. 

The central themes for this symposium were (1) social and physical
embeddedness of robots; (2) goal sharing and transfer between robots and
humans; (3) nonverbal and verbal methods for establishing and maintaining the
user's engagement; (4) factors relevant to the acceptance of assistive robots
by a community of users (especially those with special needs); and (5)
analytical frameworks and methods that can be applied to building and
evaluating socially assistive robots. 

The symposium featured 12 presentations, 10 posters, two panel presentations,
a joint session with the Interaction Challenges for Artificial Assistants
Symposium, and much discussion. The presented papers ranged from descriptions
of social capabilities needed for robots to assist humans in physical or
cognitive tasks (such as rehabilitation and training, therapeutic and
educational play, mobility) to social factors and technical designs. The
different applications for children with autism, poststroke patients, the
elderly, and the visually impaired spanned a variety of technologies and
robotic platforms. 

The presentations identified a number of relevant issues: the role of the
robot's physical embodiment; the integration of a priori knowledge about
users; ensuring safety in interaction design; the relative roles played by
social and physical robot assistance and how they can best be balanced; the
representation of a robot's perception and competences in forms accessible
and understandable to a non-technical user; and models from psychology,
cognitive science, and social science that can be utilized to advance the
goals of social assistive robotics. Furthermore, in order to promote and
facilitate interdisciplinary communication and discussion, symposium
participants took part in two hands-on exercises where they discussed,
sketched, and prototyped technologies and scenarios for assistive robots
meant for different applications. We considered issues such as intent
recognition, modality selection, empathy, adaptive systems, verbal and
nonverbal communication, and proactive assistance. 

Three invited talks were presented at the symposium. Sal Restivo (Rensselaer
Polytechnic Institute, USA) discussed principles from social theory relevant
to the development of socially assistive robots. Brian Scassellati (Yale
University, USA) discussed the use of social robots for the diagnosis,
treatment, and understanding of autism. Finally, Rachid Alami (LAAS Toulouse,
France) described work on service robots that fit in and are accepted by
humans in everyday environments. Symposium participants also had the chance
to visit two laboratories at Stanford University: Clifford Nass's
Communication between Humans and Interactive Media laboratory, which explores
the fundamental relationships between humans and interactive media; and
Oussama Khatib's robotics laboratory, which focuses on safety in humanrobot
interaction with manipulator robotics. 

Because of the novelty of this research area and high interest by
participants, there are plans to continue activities in this field at future
AAAI spring symposia and other venues. 

The papers from this symposium were published in the AAAI technical report
series and are available from AAAI Press as AAAI Technical Report SS-07-05.