Developing haptic and visual perceptual categories for reaching and grasping with a humanoid robot

Properties of the human embodiment - sensorimotor apparatus and neurologica l structure - participate directly in the growth and development of cogniti ve processes against enormous worst case complexity. It is our position tha t relationships between morphology and perception. overtime lead to increas ingly comprehensive models that describe the agent's relationship to the wo rld. We are applying insight derived from neuroscience, neurology, and deve lopmental psychology to the design of advanced robot architectures. To inve stigate developmental processes, we have begun to approximate the human sen sorimotor configuration and to engage sensory and motor subsystems in devel opmental sequences. Many such sequences have been documented in studies of infant development, so we intend to bootstrap cognitive structures in robot s by emulating some of these growth processes that bear an essential resemb lance to the human morphology. In this paper, we will show two related exam ples in which a humanoid robot determines the models and representations th at govern its behavior. The first is a model that captures the dynamics of a haptic exploration of an object with a dextrous robot hand that supports skillful grasping. The second example constructs constellations of visual f eatures to predict relative hand/object postures that lead reliably to hapt ic utility. The result is a first step in a trajectory toward associative v isual-haptic categories that bounds the incremental complexity of each stag e of development. (C) 2001 Published by Elsevier Science B.V.