An integrated approach to achieve dextrous grasping from task level specification

This paper deals with the automation of dextrous grasping in a partly known environment using a stereo vision system and a multiingered hand mounted on a robot arm. EEective grasping requires a combination of sensing and planning capabilities: sensing to construct a well-adapted model of the situation and to guide the execution of the task, and planning to determine an appropriate grasping strategy and to generate safe, feasible ma-nipulator motions. We propose an integrated approach that combines computer vision, path planning, and ma-nipulator control in three complementary activities: the reconstruction of task-oriented models of the workspace, the determination of appropriate grasping conngurations from computed`preshapes' of the hand, and the automatic generation and execution of hand/arm motions using a hybrid geometric path planner and a hybrid control system. This paper outlines the architecture of our system , discusses the techniques we have developed on grasp planning, and nishes with a brief description of work-in-progress on the implementation and some preliminary experimental results.