cs.ROApr 1, 2026

SoftHand Model-W: A 3D-Printed, Anthropomorphic, Underactuated Robot Hand with Integrated Wrist and Carpal Tunnel

Authors: Dhillon B. Merritt, Christopher J. Ford, Haoran Li, Malia Smith, Zhixing Chen, Efi Psomopoulou, Nathan F. Lepora

Organizations: All authors are with the School of Engineering Mathematics and Technology, and Bristol Robotics Laboratory, University of Bristol, Bristol, BS8 1UB, U.K.

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

This paper presents the SoftHand Model-W: a 3D-printed, underactuated, anthropomorphic robot hand based on the Pisa/IIT SoftHand, with an integrated antagonistic tendon mechanism and 2 degree-of-freedom tendon-driven wrist. These four degrees-of-acuation provide active flexion and extension to the five fingers, and active flexion/extension and radial/ulnar deviation of the palm through the wrist, while preserving the synergistic and self-adaptive features of such SoftHands. A carpal tunnel-inspired tendon routing allows remote motor placement in the forearm, reducing distal inertia and maintaining a compact form factor. The SoftHand-W is mounted on a 6-axis robot arm and tested with two reorientation tasks requiring coordination between the hand and arm's pose: cube stacking and in-plane disc rotation. Results comparing task time, arm joint travel, and configuration changes with and without wrist actuation show that adding the wrist reduces compensatory and reconfiguration movements of the arm for a quicker task-completion time. Moreover, the wrist enables pick-and-place operations that would be impossible otherwise. Overall, the SoftHand Model-W demonstrates how proximal degrees of freedom are key to achieving versatile, human-like manipulation in real world robotic applications, with a compact design enabling deployment in research and assistive settings.

Figures & tables

Fig. 1: Range of motion of the SoftHand Model-W in the ulnar (A) and radial (B) deviations, and in flexion (C) and extension (D). Corresponding degrees of freedom of the human wrist (E) are shown below (based on [ 12 ] ).
Fig. 2: Overview of the SoftHand-W. A) Technical drawing showing the palmer side with finger, palm, and write dimensions in mm\mathrm{m}\mathrm{m} . B) CAD model of a finger (palmer side up), with flexor (red) and extensor (blue) tendons.
Fig. 3: Wrist mechanism and tendon routing. A) Tendon routing through the palm and into the wrist (routing on the back of the hand is similar). B) CAD model of the 2-DoF wrist mechanism, showing the internal bearings and axles as well as palm and forearm connectors.
joint, iiaia_{i} ( mm\mathrm{m}\mathrm{m} )αi\alpha_{i} ( rad\mathrm{r}\mathrm{a}\mathrm{d} )did_{i} ( mm\mathrm{m}\mathrm{m} )θi\theta_{i} ( rad\mathrm{r}\mathrm{a}\mathrm{d} )
134π/2\pi/20θ1\theta_{1}
24800θ2\theta_{2}
TABLE I: Distal Denavit-Hartenberg parameters of the wrist.
Fig. 4: CAD model of the ‘forearm’ - an actuation box containing four Feetech STS3215 servos and tendon spools, which enable finger flexion, finger extension, radial/ulnar deviation, and flexion/extension of the wrist. Mounting for the UR5 arm is achieved via a tool flange adapter at the base.
Fig. 5: Experimental stimuli
Fig. 6: Snapshots of the rotation test under the no wrist condition. The hand lowers onto the disc (A), before grasping (B) and rotating (C). At this point the arm cannot rotate the disc any farther, so releases the disc and returns to position B, before grasping the disc and rotating to C again.
Fig. 7: Robot arm joint angles during the object rotation task, relative to initial positions. The configuration changes occur when the arm can not rotate the end-effector any farther, so must release the disc, untwist and grasp again. Wrist actuation results in a faster time to complete the task as only one configuration change is necessary (see also Fig. 6 ).
Fig. 8: The reachable workspace of the centre of the palm, when the base of the wrist is fixed.
Fig. 9: Results from the object stacking task. Difference in absolute joint angles of the UR5 relative to the initial positions, between the fixed wrist and actuated wrist conditions. Positive values mean a joint moved more without wrist actuation. Negative values mean it moved more with wrist actuation. Note that large values in joint 6 (shown in yellow) are not important for this comparison, as it relates to rotation of the end effector and does not contribute to awkward arm configurations.
Fig. 10: Pose of the end-effector during the stacking test. Cubes are picked from the green area and stacked in the purple area. Without wrist actuation, an extreme compensatory movement is necessary for the arm to pitch the cube upward, resulting in a failed placement of cube 5.
Task / CapabilityNo WristWith Wrist
General Range±30∘\pm 30^{\circ} U/R Dev, ±90∘\pm 90^{\circ} F/E
Rotation Task  
Config Changes21
Completion Time66s47s
Stacking Task  
Success (Stack)5/66/6
TABLE II: Summary of SoftHand-W capabilities and results.

Explore similar work

CardsList

  1. MM-Hand: A 21-DOF Multi-modal Modular Dexterous Robotic Hand with Remote Actuation

    Apr 19, 2026Zhuoheng Li, Qingquan Lin, Checheng Yu +5Robotic Hand DesignDexterous Manipulation

    The University of Hong Kong · City University of Hong Kong · XiFengGu Ltd.

  2. Ruka-v2: Tendon Driven Open-Source Dexterous Hand with Wrist and Abduction for Robot Learning

    Mar 27, 2026Xinqi Lucas Liu, Ruoxi Hu, Alejandro Ojeda Olarte +6Robotic Hand DesignRobotic Manipulation

    New York University Shanghai · New York University

  3. H-PAC Hand: Control-Oriented Modeling and Tendon-Elasticity Compensation for an Underactuated Robotic Hand

    Aug 17, 2026Teng Yan, Jiongxu Chen, Teng Wang +5Robotic Hand DesignTendon-Driven Actuation

    The Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China · Southeast University, Nanjing, China