Information Technology Reference
In-Depth Information
Breazeal argued (Breazeal
2002
) that it is still very difficult to develop a robot
that behaves in a naturalistic manner similar to an adult. As such, and to take advan-
tage of human social expectations, she created Kismet, a robot that behaves and is
successfully perceived as an infant even by adults.
Leonardo (Brooks et al.
2004
) is another robot specifically designed for social
interaction by means of facial expressions and life-like body poses. A social game
that uses this robot along with speech and gesture recognition, was created. The
objectives of the game were to teach the robot names and locations of different
buttons placed in front of it, and then check to see if it knew the names by asking
him to push the buttons again. Leonardo is constantly shifting its gaze between the
object and the human to direct the human's attention to what it needs help with. The
authors claim that these kinds of behaviour ensure that Leonardo acts as expected by
a socially-aware play partner.
A social robot developed by Phillips Research, the iCat robot (van Breemen
2004
),
can communicate information through multicolour LEDs in its feet and ears, can use
natural language synthesis through its speakers, and is also capable of mechanically
rendering facial expressions and give emotional feedback to the user. The iCat can be
considered a social robot since it has many of the characteristics needed to simulate
human-to-human interaction. This robot has been used to study the influence of many
social aspects such as personality (van Breemen et al.
2005
), emotional exchange
(Leite et al.
2008
) and social acceptance by older people (Heerink et al.
2008
).
Within 3D and virtual reality communities a large number of works also studies
the incorporation of socially intelligent virtual characters into virtual and augmented
reality environments (Holz et al.
2009
). However, like most studies conducted on
social robots, they are mainly focused on interactions of a single user with a single
character. Autonomous characters or robots generally lack the necessary social skills
to interact in a group. One example that tackles this challenge is a multi-agent
collaborative game called Perfect Circle (Prada and Paiva
2009
) where the user
controls a character that can interact with four other autonomous agents. The group
formed by the player and the autonomous agents must search a virtual world for
a magical item that enables them to complete the game. The autonomous agents
are endowed with social skills that allow them to interact in groups with human
members. In this game, the autonomous characters exhibit behaviours that depended
and are in agreement with the group's composition, context and structure. To win in
this simulation, players have to be aware of the social relations with the autonomous
characters. Results in this study showed that the model had positive effects on users'
social engagement, namely, on their trust and identification with the group.
These examples are somewhat successful in socially engaging users in short
term interactions. Socially engaging users in long term interactions is a much more
challenging task that is beginning to be researched in human-agent interaction.
Long Term Interactions
One of the first long-term experiments with social robots
was performed using a service robot named CERO (Huttenrauch and Eklundh
2002
).
This robot assisted motion impaired people in an office environment. After partici-
pants fully integrated the robot into their work routine, researchers concluded that
