Embodied Artificial Intelligence

On the role of morphology and materials in the emergence of cognition

Rolf Pfeifer
Artificial Intelligence Laboratory
Department of Information Technology
University of Zurich
Winterthurerstrasse 190, CH-8057 Zurich, Switzerland

This is the abstract of a talk prepared for the International interdisciplinary seminar on new robotics, evolution and embodied cognition (IISREEC).12th to 15th November 2002, Lisbon, Portugal

Abstract: Recent advances in artificial intelligence, robotics, and cognitive science have demonstrated the importance of embodiment, i.e. the idea that intelligence requires a body and cannot be understood at the level of algorithms only. There are two main types of implications of embodiment, physical and information theoretic. The former are concerned with physical forces, inertia, friction, vibrations, and energy dissipation, i.e. anything concerned with the (physical) dynamics of the system, the latter with the relation between sensory signals, motor control and neural substrate. Rather than focusing on the neural substrate only, the focus is now on the complete organism which includes morphology (shape, distribution and physical characteristics of sensors and actuators, limbs, etc.) and materials. Often, given a particular task environment, if the morphology and the materials are right, the amount of neural processing (or more generally, control) required may be dramatically reduced. An important set of new research issues concerns the understanding of “ecological balance”, i.e. the relation between morphology, materials and control.

In the talk the basic concepts are introduced and it is demonstrated how these issues play an essential role in the development of cognition. It is also shown how artificial evolution and morphogenesis can be employed for systematically investigating “ecological balance”. This requires processes of co-evolution of morphology and neural substrate that, in our approach, are modeled explicitly based on recent insights in developmental biology, in particular genetic regulatory networks. The ontogenetic development is embedded into an evolutionary cycle. These approaches yield a number of surprising results and insights that will be presented in the lecture.

More details can be found in:

Pfeifer, R., and Scheier, C. (1999). Understanding intelligence. Cambridge, Mass.: MIT Press.

Pfeifer, R. (2000). On the role of morphology and materials in adaptive behavior. In: J.-A. Meyer, A. Berthoz, D. Floreano, H. Roitblat, and S.W. Wilson (eds.). From animals to animats 6. Proc. of the 6th Int. Conf. on Simulation of Adaptive Behavior. Cambridge, Mass.: MIT Press, 23-32.

Bongard, J., and Pfeifer, R. (2002). Evolving complete agents using artificial ontogeny. To appear in: F. Hara and R. Pfeifer (eds.). Shaping embodied intelligence – the morpho-functional machine perspective. Springer Verlag.

Bongard, J., and Pfeifer, R. (2001). Repeated structure and dissociation of genotypic and phenotypic complexity in artificial ontogeny. Genetic and Evolutionary Computation Conference, GECCO-2001.

Related Links

For more information contact Luis Rocha at rocha@lanl.gov
Last Modified: October 31, 2002