Dynamic use of tactile afferent signals in control of dexterous manipulation

Abstract

During object manipulation, humans select and activate neural action programs acquired during ontogenetic development. A basic issue in understanding the control of dexterous manipulation is to learn how people use sensory information to adapt the output of these neural programs such that the fingertip actions matches the requirements imposed by the physical properties of the manipulated object, e.g., weight (mass), slipperiness, shape, and mass distribution. Although visually based identification processes contribute to predictions of required fingertip actions, the digital tactile sensors provide critical information for the control of fingertip forces. The present account deals with the tactile afferent signals from the digits during manipulation and focuses on some specific issues that the neural controller has to deal with to make use of tactile information.