Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Learning Robust Grasping Strategy Through Tactile Sensing and Adaption Skill

Yueming Hu, Mengde Li, Songhua Yang, Xuetao Li, Sheng Liu, Miao Li

TL;DR

This paper introduces an human-demonstration-based adaptive grasping policy base on tactile, which aims to achieve robust gripping while resisting disturbances to maintain grasp stability and exhibits excellent generalization ability.

Abstract

Robust grasping represents an essential task in robotics, necessitating tactile feedback and reactive grasping adjustments for robust grasping of objects. Previous research has extensively combined tactile sensing with grasping, primarily relying on rule-based approaches, frequently neglecting post-grasping difficulties such as external disruptions or inherent uncertainties of the object's physics and geometry. To address these limitations, this paper introduces an human-demonstration-based adaptive grasping policy base on tactile, which aims to achieve robust gripping while resisting disturbances to maintain grasp stability. Our trained model generalizes to daily objects with seven different sizes, shapes, and textures. Experimental results demonstrate that our method performs well in dynamic and force interaction tasks and exhibits excellent generalization ability.

Learning Robust Grasping Strategy Through Tactile Sensing and Adaption Skill

TL;DR

This paper introduces an human-demonstration-based adaptive grasping policy base on tactile, which aims to achieve robust gripping while resisting disturbances to maintain grasp stability and exhibits excellent generalization ability.

Abstract

Robust grasping represents an essential task in robotics, necessitating tactile feedback and reactive grasping adjustments for robust grasping of objects. Previous research has extensively combined tactile sensing with grasping, primarily relying on rule-based approaches, frequently neglecting post-grasping difficulties such as external disruptions or inherent uncertainties of the object's physics and geometry. To address these limitations, this paper introduces an human-demonstration-based adaptive grasping policy base on tactile, which aims to achieve robust gripping while resisting disturbances to maintain grasp stability. Our trained model generalizes to daily objects with seven different sizes, shapes, and textures. Experimental results demonstrate that our method performs well in dynamic and force interaction tasks and exhibits excellent generalization ability.

Paper Structure

This paper contains 12 sections, 10 equations, 7 figures, 1 table.

Table of Contents

  1. INTRODUCTION
  2. System Design and Implementation
  3. Initial Grasp Generator
  4. Stability Estimator
  5. Grasp Adaption
  6. Experiments
  7. Experiment Setup
  8. Data Collection
  9. Model Training
  10. Grasp Operation on Novel Objects
  11. Conclusion
  12. Acknowledgment

Figures (7)

  • Figure 2: Human demonstration and Flexible tactile sensor. A human controls the gripper manually to change the parameter $\theta$ as shown in the figure, with with adding water as an external disturbance. This figure details the mechanical structure and tactile visualization of the tactile sensor.
  • Figure 3: Schematic diagram about the self-attention mechanism for learning the adaptive grasping skills.
  • Figure 4: The experimental setup for adaptive grasping demonstrations. Both fingers of the electric parallel gripper are equipped with tactile sensors. We utilized a printed circuit board to receive tactile signals through a Bluetooth, forwarding them to a laptop. The gripper interacts with the laptop through the RS485 port and transmits its angle to the host computer through the USB port.
  • Figure 5: Expert demonstrations. Experts demonstrate adaptive grasping with different forces and controlled gripper angles, thus recording the corresponding tactile readings and angles.
  • Figure 6: Objects for testing: pill box, tea can, mouthwash, milk bottle, wine bottle, perfume, ink.
  • ...and 2 more figures