| |
Abstract:
(Invited talks)
One of the main problems in designing mobile robots is that
their behavior is the emergent result of the dynamical interaction
between the robot and the environment. The robot and the
environment can be described as a dynamical system because the
sensory state of the robot at any given time is a function of both
the environment and of the robot's previous actions. The fact that
behavior is an emergent property of the interaction between the
robot and the environment has the nice consequence that simple
robots can produce complex behaviors. However it also implies that
the properties of the emergent behavior cannot easily be predicted
or inferred from a knowledge of the rules governing the
interactions. The reverse is also true: it is difficult to predict
which rules will produce a given behavior, since behavior is the
emergent result of the dynamical interaction between the robot and
the environment. This problem can be overcome by viewing robots as
autonomous artificial organisms that develop their own skills in
close interaction with the environment through an automatic process
based on artificial evolution. This scheme, by relying on an
evaluation of the whole behavior of individuals, releases the
designer from the burden of identifying the rules that will result
in the desired behavior and allows the exploitation of emergent
forms of behavior (i.e. behaviors which heavily rely on the
interaction between the robot and the environment). I will
demonstrate this claim with a set of examples. Moreover, I will
describe the state of the art in this field, the key challenges,
and some promising directions.
Stefano Nolfi is a researcher at the Institute of Psychology,
National Research Council (CNR), Rome, and Professor of
"Educational Technologies" at the University of L'Aquila, Italy.
His research interests are in the field of neurocomputational
studies of adaptive behavior in natural and artificial agents. The
main themes underlying his work are: (a) that behavioral strategies
and neural mechanisms are understood better when an organism
(living or artificial) is caught in the act, that is when one
considers situated and embodied agents in their interaction with
the environment; (b) that to understand how natural agents behave
and to build useful artificial agents one should study how living
organisms change, phylogenetically and ontogenetically as they
adapt to their environment. In the last few years Nolfi has been
intensely involved in Evolutionary Robotics (i.e. in the attempt to
develop autonomous robots through a self-organizing process based
on artificial evolution). Together with Dario Floreano he has just
finished writing a book on this topic that will be published in the
year 2000 by MIT Press/Bradford Book.
|
