IGRASP

Abstract

As part of its “Roadmap to Useful Robotics”, Shadow identified the need for a disruptive autonomous grasping solution. The goal is to be able to grasp any object with a simple high-level command. For the solution to be robust, it relies heavily on a tight integration between the hand, or fingertip sensors, and the grasp control algorithms. In this project we will focus on developing the miniaturised tactile and proximity sensing capabilities needed for a low-cost and robust way of closing the loop in our grasping pipeline. We will focus on a system that keeps learning the best way to grasp different objects; starting from a good grasp configuration and refining it autonomously each time an object is grasped.

Lead Participant , Participant

Project Cost

Grant Offer

The Shadow Robot Company Limited, London, UNITED KINGDOM £258,461 £ 180,923

Participant

Queen Mary, University of London, United Kingdom £97,279 £ 97,279

Publications

10 25 50