BiDexHand: Design and Evaluation of an Open-Source 16-DoF Biomimetic Dexterous Hand

TL;DR

BiDexHand tackles the challenge of achieving human-like dexterity with an open-source, affordable 16-DoF cable-driven hand featuring novel phalanx designs for biomimetic finger motion. A ROS2-based control framework enables multiple input modes, including vision-based teleoperation, and a 3D-printed anti-parallelogram phalanx provides passive DIP-PIP coupling to emulate human finger kinematics. Experimental evaluation using the GRASP Taxonomy and Kapandji thumb opposability demonstrates broad dexterity (33/33 grasps) and substantial thumb mobility (9/11), alongside a fingertip force of 2.14 N and a 10 lb lifting test, underscoring manipulation versatility. By releasing open designs and software, the work aims to democratize access to advanced manipulation platforms and enable rapid iteration and deployment, with future plans for closed-loop control and learning-based grasp strategies.

Abstract

Achieving human-level dexterity in robotic hands remains a fundamental challenge for enabling versatile manipulation across diverse applications. This extended abstract presents BiDexHand, a cable-driven biomimetic robotic hand that combines human-like dexterity with accessible and efficient mechanical design. The robotic hand features 16 independently actuated degrees of freedom and 5 mechanically coupled joints through novel phalange designs that replicate natural finger motion. Performance validation demonstrated success across all 33 grasp types in the GRASP Taxonomy, 9 of 11 positions in the Kapandji thumb opposition test, a measured fingertip force of 2.14\,N, and the capability to lift a 10\,lb weight. As an open-source platform supporting multiple control modes including vision-based teleoperation, BiDexHand aims to democratize access to advanced manipulation capabilities for the broader robotics research community.

Figures (7)

• Figure 1: Overview of the BiDexHand. a) Integration of BiDexHand with the Franka FER Robot arm. b) Configuration of the BiDexHand includes four 3-DoF robotic fingers and a 4-DoF robotic thumb. c) Palmer view of the BiDexHand. The proposed design includes a coupling for Franka FER Robot Arm, and a RealSense D405 Wrist Camera.
• Figure 2: Configuration of the BiDexHand phalanx. a) Configuration of the robotic hand phalanx. b) Exploded view of the robotic phalanx.
• Figure 3: Configuration of PIP and DIP joints. a) Comparison of DvPT between the anti-parallelogram linkage and an analytical fit of the human finger van2015analytical. b) Anti-parallelogram linkage on the BiDexHand’s PIP and DIP joints.
• Figure 4: Overview of the cable-driven mechanism on BiDexHand. a) Tensioning mechanism in the servo sleeve region. b) Cable routing configuration for the robot finger’s PIP joint.
• Figure 5: Assessing the GRASP taxonomy on the BiDexHand. The robotic hand achieves 33 out of 33 grasp poses.