Collaborative Research: Sensory Integration and Sensorimotor Transformations for Dexterous Manipulation

Project: Research project


Skilled manipulation is one of the most complex human behaviors as it requires fine coordination between digit placement (kinematics) and forces (kinetics). Much of our understanding of the neural control of object manipulation is based on studies of fingertip force coordination over the last quarter century. These studies have led to the theoretical framework that sensory feedback is used to form internal representations of object dynamics through which the central nervous system (CNS) can predict the digit forces necessary for manipulation. However, the vast majority of grasp studies have constrained digit placement to given points on the object. This is a major limitation because these scenarios allow subjects to create internal representation of forces, but only at the contact points the subjects have experienced. However, in more natural grasp tasks, subjects often modulate digit placement on the object and scale fingertip forces accordingly on a trial-to-trial basis such that the manipulation goal is attained. This force modulation to digit placement must take into account sensory feedback about the object and the current state of the effectors, i.e., digit placement and forces. The mechanisms by which subjects do so to achieve high-level representations of the task goal are not known, and this represents a major gap in our understanding of how the CNS learn, plans, and execute complex motor behaviors. To address this gap, a novel approach has been developed that removes the constraint of placing the digits at predetermined locations, thereby quantifying, for the first time, the trial-to-trial coordination between digit positions and forces in a more ecologically valid fashion. The overall aim of this collaborative research proposal is to delineate the mechanisms underlying internal representations that arise from integrating sensory feedback of digit positions and forces with motor commands responsible for coordinating these two variables. Specifically, the aims are (1) to quantify the mechanisms underlying sensorimotor integration during dexterous manipulation, and (2) to determine the interaction between sensorimotor memories of past manipulations and current object visual properties underlying action planning. The overarching hypothesis is that digit placement and forces are associated with independent internal representations. This hypothesis will be tested by selectively removing tactile feedback from the fingertips, manipulating visual feedback of performance error, varying visual geometric and density cues, and through object rotation tasks that create a discrepancy between visual cues and sensorimotor memories from prior manipulations.
Effective start/end date3/15/122/28/16


  • National Science Foundation (NSF): $320,001.00


Sensory feedback
Data storage equipment
Point contacts