Advances in Tactile Sensing and Robotic Manipulation

The field of robotic manipulation is moving towards increased use of tactile sensing and multi-modal perception to improve performance in tasks such as grasping and manipulation of deformable objects. Researchers are developing new tactile sensing technologies, such as active acoustic sensing and high-resolution omnidirectional tactile sensors, to provide more accurate and robust state estimation. Additionally, there is a growing trend towards using machine learning and optimization techniques, such as reinforcement learning and multi-objective optimization, to improve the design and control of robotic systems. Notable papers include:

  • VibeCheck, which demonstrates the use of active acoustic sensing for contact-rich manipulation, and
  • PP-Tac, which presents a robotic system for picking up paper-like objects using tactile feedback in dexterous robotic hands.

Sources

A Modularized Design Approach for GelSight Family of Vision-based Tactile Sensors

Enhanced Data-driven Topology Design Methodology with Multi-level Mesh and Correlation-based Mutation for Stress-related Multi-objective Optimization

VibeCheck: Using Active Acoustic Tactile Sensing for Contact-Rich Manipulation

Grasping Deformable Objects via Reinforcement Learning with Cross-Modal Attention to Visuo-Tactile Inputs

Efficient Design of Compliant Mechanisms Using Multi-Objective Optimization

PP-Tac: Paper Picking Using Tactile Feedback in Dexterous Robotic Hands

Built with on top of