Table of Contents
Fetching ...

GEX: Democratizing Dexterity with Fully-Actuated Dexterous Hand and Exoskeleton Glove

Yunlong Dong, Xing Liu, Jun Wan, Zelin Deng

TL;DR

GEX presents a low-cost, fully actuated dexterous manipulation system that blends the GX11 three-finger hand with the EX12 exoskeleton glove to enable high-fidelity teleoperation and data collection. The hardware uses 3D-printed ABS parts with independent joint actuation across 23 DoF, achieving precise bidirectional kinematics and state observability. The software stack includes an OpenRB-150-based control SDK, Dex-Retargeting for mapping human motion to the robotic hand, and force feedback to close the control loop, all openly released. Experimental tasks demonstrate grasping, reorientation, and twisting, illustrating robust dexterous manipulation and facilitating broader research in embodied AI and skill transfer. Overall, GEX lowers the barrier to entry for dexterous robotics, offering an open, scalable platform for learning and experimentation.

Abstract

This paper introduces GEX, an innovative low-cost dexterous manipulation system that combines the GX11 tri-finger anthropomorphic hand (11 DoF) with the EX12 tri-finger exoskeleton glove (12 DoF), forming a closed-loop teleoperation framework through kinematic retargeting for high-fidelity control. Both components employ modular 3D-printed finger designs, achieving ultra-low manufacturing costs while maintaining full actuation capabilities. Departing from conventional tendon-driven or underactuated approaches, our electromechanical system integrates independent joint motors across all 23 DoF, ensuring complete state observability and accurate kinematic modeling. This full-actuation architecture enables precise bidirectional kinematic calculations, substantially enhancing kinematic retargeting fidelity between the exoskeleton and robotic hand. The proposed system bridges the cost-performance gap in dexterous manipulation research, providing an accessible platform for acquiring high-quality demonstration data to advance embodied AI and dexterous robotic skill transfer learning.

GEX: Democratizing Dexterity with Fully-Actuated Dexterous Hand and Exoskeleton Glove

TL;DR

GEX presents a low-cost, fully actuated dexterous manipulation system that blends the GX11 three-finger hand with the EX12 exoskeleton glove to enable high-fidelity teleoperation and data collection. The hardware uses 3D-printed ABS parts with independent joint actuation across 23 DoF, achieving precise bidirectional kinematics and state observability. The software stack includes an OpenRB-150-based control SDK, Dex-Retargeting for mapping human motion to the robotic hand, and force feedback to close the control loop, all openly released. Experimental tasks demonstrate grasping, reorientation, and twisting, illustrating robust dexterous manipulation and facilitating broader research in embodied AI and skill transfer. Overall, GEX lowers the barrier to entry for dexterous robotics, offering an open, scalable platform for learning and experimentation.

Abstract

This paper introduces GEX, an innovative low-cost dexterous manipulation system that combines the GX11 tri-finger anthropomorphic hand (11 DoF) with the EX12 tri-finger exoskeleton glove (12 DoF), forming a closed-loop teleoperation framework through kinematic retargeting for high-fidelity control. Both components employ modular 3D-printed finger designs, achieving ultra-low manufacturing costs while maintaining full actuation capabilities. Departing from conventional tendon-driven or underactuated approaches, our electromechanical system integrates independent joint motors across all 23 DoF, ensuring complete state observability and accurate kinematic modeling. This full-actuation architecture enables precise bidirectional kinematic calculations, substantially enhancing kinematic retargeting fidelity between the exoskeleton and robotic hand. The proposed system bridges the cost-performance gap in dexterous manipulation research, providing an accessible platform for acquiring high-quality demonstration data to advance embodied AI and dexterous robotic skill transfer learning.

Paper Structure

This paper contains 11 sections, 1 equation, 16 figures.

Figures (16)

  • Figure 1: The joint configuration of the GX11 hand.
  • Figure 2: The physical diagram of the GX11 hand.
  • Figure 3: Thumb finger
  • Figure 4: Index finger
  • Figure 5: Middle finger
  • ...and 11 more figures