pith. sign in

arxiv: 2606.24377 · v1 · pith:Z57LIMMJnew · submitted 2026-06-23 · 💻 cs.RO · cs.AR

PDS Joint: A Parametric Double-Spiral Joint Tailored for Dexterous Hands

Haoyang Li , Yibo Wen , Yixiang Fan , Yiheng Xu , Yufeng Yue This is my paper

Pith reviewed 2026-06-26 00:05 UTC · model grok-4.3

classification 💻 cs.RO cs.AR
keywords compliant jointdexterous handdirectional stiffnessspiral jointproprioceptionparametric designinductive sensingrobot hand
0
0 comments X

The pith

The PDS joint is a parametric double-spiral compliant joint that systematically shapes directional stiffness across flexion, abduction, and pronation modes in dexterous hands while enabling accurate proprioception via learned inductive sens

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a new compliant joint design for robotic hands that can be parametrically tuned to exhibit different stiffness levels depending on the direction of deformation. This addresses the challenge of creating joints that allow large, human-like motions but also provide stability for grasping and resistance to overextension. By combining two types of spiral curves with an adjustable asymmetry parameter, the joint's behavior can be shaped for specific hand functions. The design is paired with embedded sensors and a machine learning method to accurately determine the joint angle from signals, cutting error by over 40% compared to traditional fitting in the hardest direction. Such joints could make robot hands safer and more capable for everyday tasks and human interaction.

Core claim

The PDS joint enables systematic shaping of directional stiffness across multiple deformation modes including flexion/extension, abduction/adduction, and pronation/supination through the use of Archimedean and logarithmic spiral templates combined with an asymmetry ratio parameter. Experiments show non-monotonic dependence of lateral support on asymmetry, and a learned MLP mapping for inductive proprioception reduces estimation error by 41.6% versus curve fitting for abduction/adduction motion. The joints are demonstrated in an open-source dexterous hand performing grasps and contact-rich interactions.

What carries the argument

The parametric double-spiral (PDS) joint structure, which uses spiral templates and an asymmetry ratio to control stiffness distributions in multiple deformation modes.

If this is right

  • The joint allows tailoring for grasp stability and hyperextension resistance via the asymmetry parameter.
  • Stiffness landscapes vary with geometric parameters, requiring principled tuning due to non-monotonic effects.
  • Inductive sensors co-designed with the joint provide reliable state estimation under large deformation when calibrated with MLP.
  • The design integrates into full dexterous hands for object grasping and safe human interactions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The spiral-based parametric approach may reduce reliance on iterative physical prototyping for compliant mechanisms.
  • Similar asymmetry tuning could be applied to other multi-mode compliant devices beyond hands.
  • The learning-based calibration pipeline might be adapted for different sensing modalities or joint geometries.
  • Non-monotonic stiffness behavior suggests that optimal parameters may need to be found through systematic search rather than simple scaling.

Load-bearing premise

That combining Archimedean and logarithmic spirals with one asymmetry ratio parameter can produce the desired non-monotonic lateral support and reliable proprioception without needing extra geometric tweaks or accounting for material effects.

What would settle it

An experiment measuring lateral support stiffness at varying asymmetry ratios that fails to show the non-monotonic dependence, or repeated trials where the MLP does not achieve the reported error reduction for abduction/adduction.

Figures

Figures reproduced from arXiv: 2606.24377 by Haoyang Li, Yibo Wen, Yiheng Xu, Yixiang Fan, Yufeng Yue.

Figure 1
Figure 1. Figure 1: (a) An overview of the PDS-joint-based dexterous hand platform. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 4
Figure 4. Figure 4: Sensor placement for different joints: DIP/PIP, MCP, and CMC. [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figure 3
Figure 3. Figure 3: Definition of the asymmetry ratio λ. (MCP) joints of the four fingers, which requires compact design, we use an Archimedean spiral ( [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 6
Figure 6. Figure 6: Exploded view of the hand assembly. c) Palm design: In addition to serving as a mounting base, the geometry of the palm largely determines the grasp envelope, thumb opposition workspace, and overall manipulation capability. Inspired by human hand anatomy, the palm possesses a human-like curvature profile to support stable power grasps and to increase the effective contact area in grasps. The palm structure… view at source ↗
Figure 8
Figure 8. Figure 8: (a) Experimental setup for mechanical characterization in flex [PITH_FULL_IMAGE:figures/full_fig_p005_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: (a) Measured flexion/extension torque-angle curves with varying [PITH_FULL_IMAGE:figures/full_fig_p005_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: (a) Measured lateral stiffness versus flexion angle with varying [PITH_FULL_IMAGE:figures/full_fig_p006_10.png] view at source ↗
Figure 12
Figure 12. Figure 12: Drift angle versus cycle count in the cyclic loading test. [PITH_FULL_IMAGE:figures/full_fig_p006_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Raw signal characterization: fitted curves of raw inductive signal [PITH_FULL_IMAGE:figures/full_fig_p007_13.png] view at source ↗
Figure 16
Figure 16. Figure 16: Contact-rich manipulation demonstrations and synchronized [PITH_FULL_IMAGE:figures/full_fig_p007_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Illustration of impact absorption and safe human-involved [PITH_FULL_IMAGE:figures/full_fig_p008_17.png] view at source ↗
read the original abstract

Compliant joints can embed safety and adaptability into dexterous hands, but achieving large-stroke anthropomorphic motion while maintaining joint-specific, directiondependent stiffness and reliable proprioception remains challenging. This paper presents the PDS joint, a parametric doublespiral (PDS) compliant joint that enables systematic shaping of directional stiffness across multiple deformation modes, including flexion/extension, abduction/adduction, and pronation/supination. We instantiate the joint using Archimedean and logarithmic spiral templates for different hand joints and introduce an asymmetry ratio to tailor stiffness distributions for both grasp stability and hyperextension resistance. To make the joint practically usable under large deformation, we co-design embedded inductive proprioception and propose a learningbased calibration pipeline that maps raw inductive signals to joint states using ArUco-marker tracking. Experiments characterize the stiffness landscapes across geometric parameters and demonstrate a non-monotonic dependence of lateral support on asymmetry, indicating the importance of principled parameter tuning. For joint-state estimation in the most challenging abduction/adduction motion, a learned multilayer-perceptron (MLP) mapping reduces the error compared with conventional curve fitting by 41.6%. Finally, we integrate the proposed joints into an open-source dexterous hand as a demonstration platform, on which the hand grasps a set of nine everyday objects and performs safe, contact-rich human-involved interactions.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper introduces the PDS joint, a parametric double-spiral compliant joint for dexterous hands that uses Archimedean and logarithmic spiral templates assigned to different joints together with a single asymmetry ratio parameter. The design aims to enable systematic, direction-dependent stiffness shaping across flexion/extension, abduction/adduction, and pronation/supination while embedding inductive proprioception. A learning-based MLP calibration is proposed to map raw signals to joint states, reported to reduce error by 41.6% versus curve fitting in abduction/adduction. Stiffness landscapes are characterized experimentally, showing non-monotonic lateral support versus asymmetry, and the joints are integrated into an open-source hand for grasping and contact-rich tasks.

Significance. If the central claims hold, the work supplies a concrete parametric template for embedding tunable, multi-axis stiffness into compliant joints without additional actuators, which would be a useful addition to the soft-robotics and dexterous-hand literature. The co-design of inductive sensing with an MLP pipeline and the open-source hand platform are practical strengths that could support reproducibility and follow-on work. The reported non-monotonic dependence on the asymmetry ratio, if statistically robust, would also highlight the value of principled geometric tuning over ad-hoc adjustment.

major comments (2)
  1. [Abstract] Abstract and experimental results: the 41.6% error reduction for the MLP in abduction/adduction is presented without error bars, sample sizes, details on data exclusion criteria, or cross-validation procedure; these omissions are load-bearing for the claim that the learned mapping reliably outperforms conventional fitting under large deformation.
  2. [Design] Design and characterization sections: the central claim that a single asymmetry ratio plus the two spiral templates produces independent, systematic stiffness control across the three rotational modes rests on experimental observation of non-monotonic lateral support; no derivation or stiffness-tensor analysis is supplied showing how the geometry decouples the axes a priori, leaving open the possibility that observed independence is joint-specific rather than a general parametric principle.
minor comments (1)
  1. [Abstract] The abstract contains the compound word "doublespiral" without a hyphen or space; consistent terminology would improve readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We address the two major comments point by point below.

read point-by-point responses

  1. Referee: [Abstract] Abstract and experimental results: the 41.6% error reduction for the MLP in abduction/adduction is presented without error bars, sample sizes, details on data exclusion criteria, or cross-validation procedure; these omissions are load-bearing for the claim that the learned mapping reliably outperforms conventional fitting under large deformation.

    Authors: We agree that the presentation of the 41.6% error reduction lacks necessary statistical details. In the revised version we will report error bars, sample sizes, data exclusion criteria, and the cross-validation procedure for the MLP calibration to substantiate the claim. revision: yes

  2. Referee: [Design] Design and characterization sections: the central claim that a single asymmetry ratio plus the two spiral templates produces independent, systematic stiffness control across the three rotational modes rests on experimental observation of non-monotonic lateral support; no derivation or stiffness-tensor analysis is supplied showing how the geometry decouples the axes a priori, leaving open the possibility that observed independence is joint-specific rather than a general parametric principle.

    Authors: The manuscript relies on experimental characterization to demonstrate the effects of the parametric choices. We acknowledge that an a priori stiffness-tensor derivation is absent. We will add a concise geometric explanation in the design section describing how the spiral templates and asymmetry ratio influence directional stiffness, while clarifying that the observed decoupling is supported by the multi-parameter experiments rather than claimed as fully general without further analysis. revision: partial

Circularity Check

0 steps flagged

No significant circularity; design and empirical validation are self-contained.

full rationale

The paper introduces a parametric double-spiral joint geometry using Archimedean and logarithmic templates plus an asymmetry ratio, then reports experimental stiffness measurements and an MLP calibration for proprioception. No load-bearing equations, derivations, or predictions are presented that reduce reported stiffness landscapes, non-monotonic behaviors, or the 41.6% error reduction to quantities defined by the asymmetry ratio or other fitted inputs. The central claims rest on geometric instantiation, fabrication, and direct measurement rather than self-definition, fitted-input predictions, or self-citation chains. This is the expected outcome for a primarily design-plus-experiment robotics paper.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The design rests on geometric parameterization of spirals and experimental mapping; no new physical entities are postulated.

free parameters (1)
  • asymmetry ratio
    Introduced to tailor stiffness distributions for grasp stability and hyperextension resistance; its value is chosen per joint type.
axioms (1)
  • domain assumption Archimedean and logarithmic spiral templates can be used to instantiate joints with controllable directional stiffness across flexion, abduction, and pronation modes
    Stated as the basis for systematic shaping of stiffness landscapes.

pith-pipeline@v0.9.1-grok · 5780 in / 1263 out tokens · 25017 ms · 2026-06-26T00:05:59.411440+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

23 extracted references

  1. [1]

    A modular, open-source 3d printed underactuated hand,

    R. R. Ma, L. U. Odhner, and A. M. Dollar, “A modular, open-source 3d printed underactuated hand,” in2013 IEEE International Conference on Robotics and Automation (ICRA), 2013, pp. 2722–2727

    2013

  2. [2]

    Design, fabrication and control of soft robots,

    D. Rus and M. T. Tolley, “Design, fabrication and control of soft robots,”Nature, vol. 521, no. 7553, pp. 467–475, 2015

    2015

  3. [3]

    Building block-based spatial topology synthesis method for large-stroke flexure hinges,

    M. Naves, D. M. Brouwer, and R. G. K. M. Aarts, “Building block-based spatial topology synthesis method for large-stroke flexure hinges,”Journal of Mechanisms and Robotics, vol. 9, no. 4, p. 041006, 2017

    2017

  4. [4]

    ARC joint: Anthropomorphic rolling contact joint with kinematically variable torsional stiffness,

    S. Kim, E. Sung, and J. Park, “ARC joint: Anthropomorphic rolling contact joint with kinematically variable torsional stiffness,”IEEE Robotics and Automation Letters, vol. 8, no. 3, pp. 1810–1817, 2023

    2023

  5. [5]

    Identification of a tetrahedral apical cell preserved within a fossilized fern fiddlehead,

    R. Cruz and A. J. Hetherington, “Identification of a tetrahedral apical cell preserved within a fossilized fern fiddlehead,”Current Biology, vol. 35, no. 2, pp. 383–390.e2, 2025

    2025

  6. [6]

    The adaptive value of young leaves being tightly folded or rolled on monocotyledons in tropical lowland rain forest: an hypothesis in two parts,

    P. J. Grubb and R. V . Jackson, “The adaptive value of young leaves being tightly folded or rolled on monocotyledons in tropical lowland rain forest: an hypothesis in two parts,”Plant Ecology, vol. 192, pp. 317–327, 2007

    2007

  7. [7]

    Double-spiral: a bioinspired pre-programmable compliant joint with multiple degrees of freedom,

    M. Jafarpour, S. Gorb, and H. Rajabi, “Double-spiral: a bioinspired pre-programmable compliant joint with multiple degrees of freedom,” Journal of the Royal Society Interface, vol. 20, p. 20220757, 2023

    2023

  8. [8]

    Double-spiral as a bio-inspired functional element in engineering design,

    M. Jafarpour, M. Aryayi, S. N. Gorb, and H. Rajabi, “Double-spiral as a bio-inspired functional element in engineering design,”Scientific Reports, vol. 14, p. 29225, 2024

    2024

  9. [9]

    Spatially distributed biomimetic compliance enables robust anthropomorphic robotic manipulation,

    K. Junge and J. Hughes, “Spatially distributed biomimetic compliance enables robust anthropomorphic robotic manipulation,”Communica- tions Engineering, vol. 4, no. 1, p. 76, 2025

    2025

  10. [10]

    Review of flexible/stretchable sensors for soft robot,

    K. Suzumori and H. Nabae, “Review of flexible/stretchable sensors for soft robot,”Journal of Robotics and Mechatronics, vol. 37, no. 1, pp. 8–12, 2025

    2025

  11. [11]

    Hybrid magnetic–inductive angular sensor with 360 ◦ range and stray-field immunity,

    B. Brajon, L. Lugani, and G. Close, “Hybrid magnetic–inductive angular sensor with 360 ◦ range and stray-field immunity,”Sensors, vol. 22, no. 6, p. 2153, 2022

    2022

  12. [12]

    Inductive sensing design guide (an219207),

    Infineon Technologies, “Inductive sensing design guide (an219207),” Infineon Technologies, Tech. Rep., 2024, application note

    2024

  13. [13]

    High-resolution low latency and low-power absolute inductive angle encoder reference design (TIDA-010961),

    Texas Instruments, “High-resolution low latency and low-power absolute inductive angle encoder reference design (TIDA-010961),” Texas Instruments, Tech. Rep., 2025, Design Guide, released Sep 23,

    2025

  14. [14]

    Available: https://www.ti.com/tool/TIDA-010961

    [Online]. Available: https://www.ti.com/tool/TIDA-010961

  15. [15]

    3d- printed flexure-based finger joints for anthropomorphic hands,

    L. A. Garcia Rodriguez, M. Naves, and D. M. Brouwer, “3d- printed flexure-based finger joints for anthropomorphic hands,” in2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2018, pp. 1437–1442

    2018

  16. [16]

    Compliant joint actuator with dual spiral springs,

    Y . Kim, J. Lee, and J. Park, “Compliant joint actuator with dual spiral springs,”IEEE/ASME Transactions on Mechatronics, vol. 18, no. 6, pp. 1839–1844, 2013

    2013

  17. [17]

    Study on the large-displacement behaviour of a spiral spring with variations of cross-section, orthotropy and prestress,

    G. Radaelli and J. L. Herder, “Study on the large-displacement behaviour of a spiral spring with variations of cross-section, orthotropy and prestress,”Mechanical Sciences, vol. 9, pp. 337–348, 2018

    2018

  18. [18]

    Minimal sensing approach of an underactuated flexure based gripper for agri-food applications,

    J. Korenblik, “Minimal sensing approach of an underactuated flexure based gripper for agri-food applications,” M.Sc. Thesis, University of Twente, Aug. 2021

    2021

  19. [19]

    Shielded soft force sensors,

    B. Aksoy, Y . Hao, G. Grasso, K. M. Digumarti, V . Cacucciolo, and H. Shea, “Shielded soft force sensors,”Nature Communications, vol. 13, no. 1, p. 4649, 2022

    2022

  20. [20]

    Inductive position sensors,

    Renesas Electronics, “Inductive position sensors,” Product page, 2026, accessed 2026-03-06. [Online]. Available: https://www.renesas. com/en/products/sensor-products/inductive-position-sensors

    2026

  21. [21]

    Automatic generation and detection of highly reliable fiducial markers under occlusion,

    S. Garrido-Jurado, R. Mu ˜noz-Salinas, F. J. Madrid-Cuevas, and M. J. Mar´ın-Jim´enez, “Automatic generation and detection of highly reliable fiducial markers under occlusion,”Pattern Recognition, vol. 47, no. 6, pp. 2280–2292, 2014

    2014

  22. [22]

    LDC1612, LDC1614: Multi-channel 28-bit induc- tance to digital converter (ldc) for inductive sensing, datasheet (rev. a),

    Texas Instruments, “LDC1612, LDC1614: Multi-channel 28-bit induc- tance to digital converter (ldc) for inductive sensing, datasheet (rev. a),” https://www.ti.com/product/LDC1614, 2018, accessed 2026-03-03

    2018

  23. [23]

    A soft inductive bi- modal sensor for proprioception and tactile sensing of soft machines,

    Y . Peng, H. Wu, Z. Wang, Y . Wang, and H. Wang, “A soft inductive bi- modal sensor for proprioception and tactile sensing of soft machines,” Soft Robotics, vol. 11, no. 6, pp. 1055–1067, 2024

    2024