Paper Image

Tactile sensing and control for dexterous in-hand manipulation

Published on:

18 January 2024

Primary Category:


Paper Authors:

Antonia Bronars,

Sangwoon Kim,

Parag Patre,

Alberto Rodriguez


Key Details

Combines tactile sensing and motion control for in-hand manipulation

Discrete pose estimator tracks objects on a grid using tactile images

Continuous estimator refines pose estimate for precise control

Achieves desired configurations and outperforms single-shot methods

Enables closed-loop tasks like assembly and tool use with limited vision

AI generated summary

Tactile sensing and control for dexterous in-hand manipulation

This paper introduces a method to estimate and control the pose of objects grasped in a robot hand, by integrating measurements from sensors on the robot and tactile sensors on the gripper fingers. It allows manipulating objects precisely, even when visual perception is limited. This enables closed-loop applications like assembly and tool use. The approach combines a coarse discrete pose estimator with a continuous estimator and motion controller for fine manipulation.

Answers from this paper


No comments yet, be the first to start the conversation...

Sign up to comment on this paper

Sign Up