pith. sign in

arxiv: 2605.01434 · v1 · submitted 2026-05-02 · 💻 cs.RO

High-Speed, Scalable Sensor Readout for Dexterous Robotic Hands via Shift-Register Multiplexing

Jaehoon Kim , Lazaros Christoforidis , Michalis Papadakis , Victor Kartsch , Robert K. Katzschmann This is my paper

Pith reviewed 2026-05-09 14:30 UTC · model grok-4.3

classification 💻 cs.RO
keywords sensor readout architecturerobotic handsshift-register multiplexingdexterous manipulationtactile sensingjoint angle sensinghigh-speed samplingscalable wiring
0
0 comments X

The pith

A serial-in parallel-out shift-register architecture lets robotic hands read 20 heterogeneous analog sensors at 1 kHz using only three wires between modules.

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

The paper introduces a multiplexing method that chains sensor modules together so each additional module needs just three shared signal lines instead of individual wires. This removes the usual trade-off between the number of sensors, wiring bulk, and sampling speed that has limited fully instrumented dexterous hands. The authors integrate the scheme into a tendon-driven hand with 16 joint modules plus a four-channel tactile module and show that 20 channels can be scanned at 1 kHz while keeping joint-angle error below one degree and force-estimation RMSE at 0.125 N. Because the same three lines also support heterogeneous sensor types and configurable scan rates up to 1.5 kHz, the architecture directly addresses the wiring and bandwidth bottlenecks that currently prevent dense, high-speed sensing on multi-fingered robots.

Core claim

The proposed SIPO shift-register readout architecture supports versatile integration of heterogeneous analog-output sensors, scalable expansion using only three signal lines between sensor modules, and fast, configurable sampling. Validation on a tendon-driven robotic hand with 16 joint sensor modules and one four-channel tactile module enables acquisition of 20 sensor channels at a full-scan rate of 1 kHz with stable operation up to 1.5 kHz. Joint sensor characterization showed a maximum slope absolute percentage error of 0.446 percent and sub-degree estimation error, while LSTM models on the tactile data reached 0.125 N RMSE for force and 93.4 percent accuracy for contact location.

What carries the argument

Serial-in parallel-out (SIPO) shift-register multiplexing that chains sensor modules with three shared lines (clock, data, and enable) so each module decodes its own channel selection and routes its analog output to a common bus.

If this is right

  • The same three-line bus can accept any mix of analog sensors, so joint potentiometers, tactile arrays, and future modalities can share one readout without custom electronics per sensor type.
  • Scan rate is configurable per module, allowing high-speed sampling on critical channels while lower-rate sensors share the bus without reducing overall throughput.
  • Full 1 kHz acquisition of 20 channels supplies joint and contact data fast enough for real-time force estimation and contact classification during manipulation.
  • System-level tests already show joint sensors outperform motor-current estimates during contact, so the architecture supplies higher-quality feedback for closed-loop control.

Where Pith is reading between the lines

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

  • Because each module only needs three wires regardless of how many channels it contains, the same bus could be extended to entire arms or multi-robot teams without exponential growth in cabling.
  • The low reported noise impact suggests the method could be combined with existing high-density tactile skins or optical sensors that currently cannot be read at kilohertz rates due to wire count.
  • Deploying the same modules on different robot platforms would require only firmware changes to the shift-register addressing, providing a reusable hardware substrate for dexterous sensing research.

Load-bearing premise

That chaining the modules through shift registers will not introduce enough electrical noise, timing skew, or signal degradation to measurably harm the accuracy of the attached analog sensors.

What would settle it

Replicate the 20-channel hand setup but replace the shift-register chain with direct parallel wiring to each sensor and compare slope APE, sub-degree joint error, and tactile force RMSE at identical 1 kHz and 1.5 kHz scan rates.

Figures

Figures reproduced from arXiv: 2605.01434 by Jaehoon Kim, Lazaros Christoforidis, Michalis Papadakis, Robert K. Katzschmann, Victor Kartsch.

Figure 1
Figure 1. Figure 1: Overview of the proposed sensorized robotic hand. (a) Fully view at source ↗
Figure 2
Figure 2. Figure 2: Operating principle of the serial-in parallel-out (SIPO) shift register view at source ↗
Figure 3
Figure 3. Figure 3: Configuration and hardware implementation of the proposed sensor view at source ↗
Figure 4
Figure 4. Figure 4: Mechanical integration of the sensor system into the robotic hand. (a) Front view of the index finger showing the integrated sensing regions. (b) view at source ↗
Figure 5
Figure 5. Figure 5: Experimental setup for characterization of the joint sensor view at source ↗
Figure 6
Figure 6. Figure 6: Characterization of the tactile sensor module. (a) CNC-based view at source ↗
Figure 7
Figure 7. Figure 7: Evaluation of the proposed sensorized robotic hand. (a) Experimen view at source ↗
read the original abstract

Dexterous robotic hands require high-speed multimodal sensing across many degrees of freedom, yet existing readout architectures often impose trade-offs between sensor count, wiring complexity, and sampling bandwidth. This paper presents a scalable analog sensor readout architecture based on a serial-in parallel-out (SIPO) shift-register principle. The proposed architecture supports versatile integration of heterogeneous analog-output sensors, scalable expansion using only three signal lines between sensor modules, and fast, configurable sampling. We validate the approach on a tendon-driven robotic hand integrating 16 joint sensor modules and one four-channel tactile sensor module, enabling acquisition of 20 sensor channels at a full-scan rate of 1 kHz, with stable operation up to 1.5 kHz. Joint sensor characterization showed a maximum slope absolute percentage error (APE) of 0.446% and sub-degree estimation error, indicating that the proposed readout system does not significantly degrade sensing performance. For tactile sensing, LSTM-based models achieved an RMSE of 0.125 N for force estimation and 93.4% accuracy for five-class contact-location classification, and were deployed for real-time inference at 1 kHz. System-level experiments showed that the joint sensors provide more accurate feedback than motor-based estimation during interaction, while the tactile sensor enables responsive force estimation in contact. The proposed architecture offers a practical path toward fully sensorized robotic hands for dexterous manipulation.

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

0 major / 2 minor

Summary. The manuscript proposes a scalable analog sensor readout architecture based on serial-in parallel-out (SIPO) shift-register multiplexing. It claims to enable integration of heterogeneous sensors using only three signal lines between modules, with configurable high-speed sampling. Validation is performed on a tendon-driven robotic hand with 16 joint sensor modules and one 4-channel tactile module, achieving 20-channel acquisition at 1 kHz full-scan rate (stable to 1.5 kHz). Reported results include maximum slope absolute percentage error (APE) of 0.446% and sub-degree estimation error for joint sensors, LSTM-based tactile force estimation with RMSE 0.125 N, 93.4% contact-location classification accuracy, and real-time 1 kHz inference, with system experiments showing improved feedback over motor-based estimation.

Significance. If the reported performance holds, the architecture provides a practical, low-wiring solution for high-speed multimodal sensing in dexterous hands, addressing key trade-offs in sensor count, bandwidth, and complexity. The work is strengthened by concrete end-to-end physical validation on a real tendon-driven hand, successful real-time ML deployment, and quantitative error metrics demonstrating minimal degradation of sensor accuracy.

minor comments (2)
  1. The abstract and introduction would benefit from explicitly stating the number of degrees of freedom and joint types in the tendon-driven hand to better contextualize the choice of 16 joint modules.
  2. Figure captions and the experimental setup description could include additional details on sensor models, exact wiring configuration, and power consumption measurements to improve reproducibility.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive summary, recognition of the practical significance of the SIPO shift-register architecture, and recommendation to accept the manuscript. No major comments were raised, so we have no points requiring response or revision.

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper is an applied hardware design and experimental validation paper. It describes a SIPO shift-register multiplexing architecture for sensor readout and validates it through physical implementation on a robotic hand, reporting concrete metrics such as 1 kHz sampling, 0.446% APE, and LSTM accuracies. No mathematical derivations, first-principles predictions, fitted parameters renamed as predictions, or self-citation chains appear in the provided text or abstract. All performance claims reduce directly to measured experimental results rather than to internal definitions or prior self-citations.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

This is an applied engineering paper describing a hardware architecture and its empirical validation on a robotic hand. No free parameters, mathematical axioms, or new invented entities are introduced; the design relies on standard digital electronics components and principles.

pith-pipeline@v0.9.0 · 5572 in / 1343 out tokens · 39142 ms · 2026-05-09T14:30:48.237238+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

16 extracted references · 16 canonical work pages

  1. [1]

    A review on sensory perception for dexterous robotic manipulation,

    Z. Xia, Z. Deng, B. Fang, Y . Yang, and F. Sun, “A review on sensory perception for dexterous robotic manipulation,”International Journal of Advanced Robotic Systems, vol. 19, no. 2, p. 17298806221095974, Mar. 2022

    work page 2022

  2. [2]

    Sensors for Robotic Hands: A Survey of State of the Art,

    A. Saudabayev and H. A. Varol, “Sensors for Robotic Hands: A Survey of State of the Art,”IEEE Access, vol. 3, pp. 1765–1782, 2015

    work page 2015

  3. [3]

    Learning dexterous in-hand manipulation,

    O. M. Andrychowicz, B. Baker, M. Chociej, R. J ´ozefowicz, B. Mc- Grew, J. Pachocki, A. Petron, M. Plappert, G. Powell, A. Ray, J. Schneider, S. Sidor, J. Tobin, P. Welinder, L. Weng, and W. Zaremba, “Learning dexterous in-hand manipulation,”The International Journal of Robotics Research, vol. 39, no. 1, pp. 3–20, Jan. 2020

    work page 2020

  4. [4]

    DexTouch: Learning to Seek and Manipulate Objects With Tactile Dexterity,

    K.-W. Lee, Y . Qin, X. Wang, and S.-C. Lim, “DexTouch: Learning to Seek and Manipulate Objects With Tactile Dexterity,”IEEE Robotics and Automation Letters, vol. 9, no. 12, pp. 10 772–10 779, Feb. 2024

    work page 2024

  5. [5]

    Tactile Dexterity: Manipulation Primitives with Tactile Feedback,

    F. R. Hogan, J. Ballester, S. Dong, and A. Rodriguez, “Tactile Dexterity: Manipulation Primitives with Tactile Feedback,” in2020 IEEE International Conference on Robotics and Automation (ICRA), May 2020, pp. 8863–8869

    work page 2020

  6. [6]

    ”FRCEF: The new friction reduced and coupling enhanced finger for the Awiwi hand

    W. Friedl, M. Chalon, J. Reinecke, and M. Grebenstein, “”FRCEF: The new friction reduced and coupling enhanced finger for the Awiwi hand”,” in2015 IEEE-RAS 15th International Conference on Humanoid Robots (Humanoids), Jan. 2015, pp. 140–147

    work page 2015

  7. [7]

    Leveraging Pre- trained Latent Representations for Few-Shot Imitation Learning on an Anthropomorphic Robotic Hand,

    D. Liconti, Y . Toshimitsu, and R. Katzschmann, “Leveraging Pre- trained Latent Representations for Few-Shot Imitation Learning on an Anthropomorphic Robotic Hand,” in2024 IEEE-RAS 23rd Inter- national Conference on Humanoid Robots (Humanoids), Jan. 2024, pp. 181–188

    work page 2024

  8. [8]

    Coding and use of tactile signals from the fingertips in object manipulation tasks,

    R. S. Johansson and J. R. Flanagan, “Coding and use of tactile signals from the fingertips in object manipulation tasks,”Nature Reviews Neuroscience, vol. 10, no. 5, pp. 345–359, May 2009

    work page 2009

  9. [9]

    Spacehand: a multi-fingered robotic hand for space,

    M. Chalon, M. Maier, W. Bertleff, A. Beyer, R. Bayer, W. Friedl, P. Neugebauer, T. Obermeier, H.-J. Sedlmayr, N. Seitz, and A. Stem- mer, “Spacehand: a multi-fingered robotic hand for space,” inASTRA 2015, Nordwick, Netherlands, 2015

    work page 2015

  10. [10]

    neoDavid: A Humanoid Robot With Variable- Stiffness Actuation and Dexterous Manipulation Skills,

    S. Wolf, T. Bahls, B. Deutschmann, A. Dietrich, M. Harder, H. H ¨oppner, C. Hofmann, A. H. Martin, M. Keppler, L. Kl ¨upfel, H. Maurenbrecher, X. Meng, A. E. Reichert, M. Stoiber, M. Bihler, M. Chalon, O. Eiberger, W. Friedl, M. Grebenstein, M. Iskandar, V . Langofer, M. Pfanne, A. Raffin, J. Reinecke, T. W ¨usthoff, and A. Albu-Sch ¨affer, “neoDavid: A H...

    work page 2025

  11. [11]

    ORCA: An Open-Source, Reliable, Cost- Effective, Anthropomorphic Robotic Hand for Uninterrupted Dexter- ous Task Learning,

    C. C. Christoph, M. Eberlein, F. Katsimalis, A. Roberti, A. Sym- petheros, M. R. V ogt, D. Liconti, C. Yang, B. G. Cangan, R. J. Hinchet, and R. K. Katzschmann, “ORCA: An Open-Source, Reliable, Cost- Effective, Anthropomorphic Robotic Hand for Uninterrupted Dexter- ous Task Learning,” in2025 IEEE/RSJ International Conference on Intelligent Robots and Syst...

    work page 2025

  12. [12]

    An introduction to I2C and SPI protocols,

    F. Leens, “An introduction to I2C and SPI protocols,”IEEE Instru- mentation & Measurement Magazine, vol. 12, no. 1, pp. 8–13, Feb. 2009

    work page 2009

  13. [13]

    Optimization based full body control for the atlas robot,

    S. Feng, E. Whitman, X. Xinjilefu, and C. G. Atkeson, “Optimization based full body control for the atlas robot,” in2014 IEEE-RAS International Conference on Humanoid Robots, Jan. 2014, pp. 120– 127

    work page 2014

  14. [14]

    Vibrotactile thresholds measured at the finger,

    R. T. Verrillo, “Vibrotactile thresholds measured at the finger,”Per- ception & Psychophysics, vol. 9, no. 4, pp. 329–330, Jul. 1971

    work page 1971

  15. [15]

    Tactile Sensor for Mea- suring Load Distribution Using Selected Scanning Method with a Few Wires,

    T. Nishino, M. Shimojo, and M. Ishikawa, “Tactile Sensor for Mea- suring Load Distribution Using Selected Scanning Method with a Few Wires,”Transactions of the Society of Instrument and Control Engineers, vol. 45, no. 8, pp. 391–397, 2009

    work page 2009

  16. [16]

    Use of Deep Learning for Characterization of Microfluidic Soft Sensors,

    S. Han, T. Kim, D. Kim, Y .-L. Park, and S. Jo, “Use of Deep Learning for Characterization of Microfluidic Soft Sensors,”IEEE Robotics and Automation Letters, vol. 3, no. 2, pp. 873–880, Apr. 2018

    work page 2018