A Humanoid Robot Hand for Dexterous Manipulation
Abstract
The Pace University Robotics Lab is developing a robotic architecture based on cognitive principles for autonomous robots that can s,afely interact and collaborate with people on a wide range of physical tasks. Central to our architecture is its mental model, which, is a 3D virtual world that the architecture synchronizes with the environment in real time. This mental model represents changes in, causation in the world and generates expectations about physical change in the world. A main scientific focus of this project is th,e use of this spatial representation in planning, learning and communication in Human-Robot interaction. This project is supported b,y the Office of Naval Research under DURIP grant N00014-19-1-2546, which supported the purchase of a mobile humanoid robot from Clea,rpath Robotics.This proposal is for funding to support the purchase of an advanced humanoid robot hand. At the time our original DUR,IP grant was funded, there was not a hand with the dexterity and sensors that could work on our humanoid robot, so we did not includ,e a request for funding for a hand. The available hands were too heavy. For example, the Shadow Hand weighs more than the humanoid r,obots Gen3 arms could lift. As a result, we went with the default grippers, which are light but are simple 2-degree of freedom grip,pers. This imposes severe limitations on the manipulation tasks that our robot can perform.The hand we have selected is the Shadow H,and Lite, which has recently been developed. It weighs only 2.4 kilograms, and has sophisticated dexterity, as well as many sensors,, including pressure, force and touch sensors. This reduction of weight is accomplished by removing the smallest (pinky) finger and t,he wrist, as well as shortening the forearm. The Gen3 arm can lift 4.4 kilograms, leaving sufficient capacity for lifting and manipu,lating objects weighing up to 2 kilograms. We are asking for funding for one hand, leaving the basic gripper on the other arm, which, will be able to lift the full 4.4 kilograms. The total lifting capability of the robot will be only slightly reduced, and its dexte,rity will be greatly enhanced.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 01, 2022
- Source ID
- N000142212285
Entities
People
- David Benjamin
Organizations
- Office of Naval Research
- Pace University
- United States Navy