A Universal Wire Testing Machine for Enhancing the Performance of Wire-Driven Robots

Kei Okada; Kento Kawaharazuka; Temma Suzuki

REVIEW 3 major objections 4 minor 19 references

A Universal Wire Testing Machine for Enhancing the Performance of Wire-Driven Robots

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A wire testing machine that measures real pulley losses cuts a wire-driven robot's vertical force error by about 13 percent.

desk verdict Useful hardware and a useful efficiency dataset; the robot demo's 13% claim needs error bars before it carries weight. read the letter →

arxiv 2509.10862 v1 pith:5N22NNOB submitted 2025-09-13 cs.RO

Temma Suzuki , Kento Kawaharazuka , Kei Okada This is my paper · ORCID

classification cs.RO

keywords wire-drivenrobotstensiontransmissionefficiencypassivepulleyforcecontrolwiretestingmachinecoupledtendon-drivenmechanismpre-stretchingvariable-lengthdynamics

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper argues that the main obstacle to wider use of wire-driven robots, namely that flexible wires are lossy and hard to model, can be addressed by measuring the wires themselves. The authors built a Universal Wire Testing Machine that pre-stretches new wires, measures per-pulley tension transmission efficiency for eight pulley diameters and four wire types, and records tension dynamics while wire length changes. They then feed the measured per-pulley efficiency into the quadratic-program tension distributor of a real wire-driven robot. With compensation, the vertical end-effector force RMSE fell from 10.9 N to 9.5 N. The point is that a practical characterization rig, not a more complex model, is enough to make wire transmission more controllable.

What carries the argument

The load-bearing object is the tension transmission efficiency matrix used inside the quadratic program. For each wire $i$ and joint $j$, the number of passive pulleys $N_{ij}$ converts the measured per-pulley efficiency $\eta_p$ into $\eta_{ij} = \eta_p^{N_{ij}}$; the Hadamard product $\eta \odot G$ scales each wire's moment-arm contribution by the fraction of tension that survives the pulley routing, so the optimizer asks for more tension on wires that have lost more. The quadratic program minimizes $(\tau_{\text{ref}} + (\eta \odot G)^T T_{\text{ref}})^T \Lambda (\tau_{\text{ref}} + (\eta \odot G)^T T_{\text{ref}}) + |T_{\text{ref}}|^2$ subject to $T_{\min} \leq T_{\text{ref}} \leq T_{\max}$, using the quadratic program formulation from prior work on redundant muscle tension. The testing machine itself supplies the measured $\eta_p$ values, along with pre-stretching to remove initial wire stretch and a variable-length wire dynamics measurement system.

What would settle it

Measure the actual tension just before and after every passive pulley on the robot while the end effector executes the same 0 to 40 N ramp, under both directions of wire motion, and compare each pulley's live efficiency with the single measured per-pulley efficiency. If live efficiencies differ substantially from the test-rig value, or if the compensation changes the sign of the tension correction on reversing wires, the 13 percent RMSE reduction should disappear or become negative.

Watch

Extended reading notes

Core claim

The central discovery is that tension lost when a wire rounds a passive pulley can be turned from unmodeled error into a compensable quantity. The paper measures per-pulley efficiency as $\eta_p = \sqrt{T_{\text{out}}/T_{\text{in}}}$, then constructs an efficiency matrix $\eta$ whose entries are $\eta_p^{N_{ij}}$, where $N_{ij}$ is the number of passive pulleys wire $i$ passes through before reaching joint $j$. Replacing the muscle-length Jacobian $G$ by the elementwise product $\eta \odot G$ in the quadratic program that solves for target wire tensions (Eqs. 5 and 7) makes the optimizer request extra tension on wires that have lost more. In a robot using 1 mm Vectran rope and 12 mm pulleys, with up to seven passive pulley units per wire, this compensation reduced the vertical end-effector force RMSE from 10.9 N to 9.5 N, about 13 percent. The same experiments show transmission efficiency increases monotonically with pulley diameter and decreases with wire diameter, and that variable-length wires have different tension dynamics than constant-length wires.

Load-bearing premise

The compensation assumes that the efficiency measured on the test rig at a 90-degree wrap angle and at 200 N and 400 N input tension applies unchanged to every pulley in the robot under actual tension, wire velocity, wrap angle, and direction of motion, even though the robot's wires reverse direction.

Editorial extensions

If this is right

Force control on existing coupled wire-driven robots can be improved by roughly 13 percent simply by replacing the lossless muscle Jacobian with the measured-efficiency version, provided per-pulley efficiencies are known.
Increasing passive pulley diameter and decreasing wire diameter raises measured transmission efficiency monotonically, giving a direct design rule for low-friction wire routing.
Variable-length wires lose about 5 dB of tension-tracking magnitude at low frequencies and shift and attenuate resonance compared with constant-length wires, so dynamic models that ignore length variation will overestimate tension tracking.
Pre-stretching synthetic fiber rope removes initial plastic stretch, as demonstrated by an 8.2 m Dyneema rope elongating 0.43 m under 510 N for 12 hours, which should make strain-tension behavior more reproducible over the robot's life.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A natural extension the paper leaves implicit is to make the efficiency matrix configuration-dependent: because wrap angles and the number of active pulley contacts change with robot posture, a lookup table indexed by joint angles could shrink the residual error further than the single 90-degree measurement.
The same per-pulley efficiency data could be inverted into a design tool: for a target force accuracy and available pulley diameters, a designer could choose wire diameter and routing to keep any wire's cumulative efficiency above a threshold.
Since prior work cited in the paper shows friction spikes when wires reverse direction, a direction-symmetric constant efficiency is the weakest point of the compensation; testing the compensated controller under cyclic reversing loads would show whether the 13 percent gain persists.
The variable-length dynamics measurement suggests a concrete testable claim: a tension-prediction model trained on length-varying data should beat a fixed-length viscoelastic model in the 2-10 Hz band where phase lag develops.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague, this paper is worth your time for the hardware and the dataset, not for the headline robot demo. The testing machine is a genuinely useful piece of kit: it pre-stretches wires, measures tension transmission efficiency across eight pulley diameters and four rope types at 200 and 400 N, and adds a variable-length dynamic response measurement that prior work didn't cover. The efficiency curves are clean and monotonic – larger pulleys lose less, thicker wires lose more – and the numbers are consistent enough to be a design reference. The variable-length frequency response showing a ~5 dB low-frequency drop is a new observation that matters for any robot where wire length changes with posture.

The soft spot is the validation experiment. The claim that compensation reduces end-effector force error by 13% rests on two mean RMSEs: 10.9 N vs 9.5 N, five trials each, with no standard deviation, confidence interval, or test. Looking at Fig. 11, the two curves overlap for most of the ramp, so the difference could easily come from one phase of the motion or from trial noise. That's a real weakness in the paper's central quantitative claim. It doesn't undermine the utility of the efficiency table, but it does mean the robot control application isn't proven yet.

A second, smaller issue is transferability: the per-pulley efficiency was measured at a fixed 90° wrap angle and two tension setpoints, then applied to a robot whose wires reverse direction and see varying tensions. The authors mention the wrap-angle limitation in the conclusion, but don't address direction reversal, and prior work (Choi et al.) shows that constant-efficiency models miss friction transients on reversal. So the compensation model is a reasonable first step, not a general method.

The paper is honest and clearly written. It cites the right prior work, reports its setup in enough detail to reproduce, and doesn't oversell beyond the data – except for the 13% claim, which is a simple overstatement given the missing error bars. I'd send this to peer review, but the referee should ask for per-trial RMSEs, variance, and ideally a wrap-angle sweep or at least a discussion of why direction reversal might be captured by the mean efficiency. It's a solid hardware contribution that deserves a serious look, and the dataset alone is likely to get cited.

Referee Report

3 major / 4 minor

Summary. The paper proposes a Universal Wire Testing Machine with three functions: pre-stretching wires to remove initial plastic elongation, measuring tension transmission efficiency for combinations of four wire types and eight pulley diameters, and measuring tension dynamics under variable total wire length. The authors report efficiency trends, characterize variable-length versus constant-length frequency responses, and apply the measured per-pulley efficiency in a quadratic-program tension allocator (Eq. (7)) for force control of a wire-driven robot. The central quantitative claim is that this compensation reduced the vertical end-effector force RMSE from 10.9 N to 9.5 N, about 13%, based on five trials per condition.

Significance. The hardware contribution is useful: the testing machine covers a wider range of pulley diameters and wire types than prior studies, and the efficiency trends—increasing with pulley diameter and decreasing with wire diameter—are physically sensible and consistent with the reported measurement curves. The dynamic variable-length wire measurements also address a real gap in the literature. However, the paper's main claim that the testing-machine data improve robot force control rests on two mean RMSE values with no reported uncertainty, and the transfer of a single per-pulley efficiency to all robot operating conditions is not established. If these issues are resolved, the work could be a solid practical contribution to wire-driven robot design and control; as it stands, its significance is contingent.

major comments (3)

[Section IV-C, Fig. 11] The claimed 13% force-error reduction is supported only by the difference between two mean RMSE values, 10.9 N without compensation and 9.5 N with compensation, with five trials per condition and no standard deviation, confidence interval, or significance test. The two curves in Fig. 11 appear to overlap substantially over the trial, so the mean difference could be dominated by a short phase of the ramp rather than a consistent reduction. Please report per-trial RMSE values with dispersion and a statistical comparison, or explicitly soften the claim to a preliminary observation.
[Section IV-A and Eq. (5)] The compensation model uses a single measured per-pulley efficiency, obtained at a fixed 90-degree wrap angle and at 200 N and 400 N input tensions, as an exponentiated efficiency matrix for every passive pulley in the robot, independent of actual tension, wrap angle, and direction of motion. Section V acknowledges the wrap-angle limitation, but the robot wires may also reverse direction, and prior work [16] shows that direction changes introduce friction not captured by constant-efficiency models. The authors should either demonstrate that the test-bench conditions are representative of the robot's operating conditions or restrict the claim accordingly; otherwise the 9.5 N versus 10.9 N comparison may not transfer beyond the specific configuration tested.
[Section IV-C] The experimental description does not state the robot posture, the range of wire tensions actually generated, or whether the five trials in each condition used identical trajectories and initial states. Without this information, it is difficult to assess whether the reported RMSE difference is attributable to the compensation term rather than to trial-specific variability or configuration differences.

minor comments (4)

[Section III-B, Fig. 4] The text reports a plastic deformation of 0.43 m, while the Fig. 4 caption states that the wire elongated from 8.2 m to 8.6 m, which is 0.4 m; please reconcile these numbers.
[Section IV-A, Figs. 7 and 8] Twenty measurements per condition were averaged to produce each efficiency point, but no error bars or dispersion measures are shown; providing them would help assess the reliability of the reported trends.
[References] The reference list is not in first-citation order (e.g., [4] appears before [2] and [3]), and the in-text citation of [16] as 'Sung-Hyun et al.' should be checked against the actual author names; standard numeric citation practice would also avoid this ambiguity.
[Section IV-B, Fig. 9] The frequency-response comparison would be clearer if the authors defined exactly what distinguishes the 'variable-length' from the 'constant-length' condition beyond the linear loading unit being free or fixed, and if they reported whether the plotted curves are from single trials or averaged over repetitions.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the efficiency data are measured independently before the robot trial, and the force-error comparison is a forward experimental validation rather than a fitted prediction.

full rationale

The paper's central claim is that its Universal Wire Testing Machine measures tension transmission efficiency and that applying that measured efficiency to force control reduces end-effector force error. The efficiency values used in the compensation model, Eq. (5), are obtained in Section IV-A from independent tension measurements before and after instrumented passive pulleys (E = sqrt(Tout/Tin)), with twenty repeated trials per condition. These values are not fitted to the robot's force-error outcome. The robot experiment in Section IV-C then applies the pre-measured per-pulley efficiency through the exponentiated matrix eta in Eq. (6) and compares the resulting force RMSE (9.5 N) with the uncompensated RMSE (10.9 N). Because the compensation gain is fixed before the robot trial and the measured error is not used to adjust eta, the 13% reduction is an empirical test of transferability, not a reduction by construction. The QP formulation in Eqs. (4) and (7) is a standard redundant-tension distribution method, cited to the authors' prior work [19] but not derived from the force error being predicted. Self-citations such as [6], [7], and [19] are contextual and do not carry the load-bearing argument. The acknowledged limitation that eta was measured at a fixed 90-degree wrap angle and only at 200 N and 400 N affects generality, not circularity. No step in the derivation chain equates a fitted input with the claimed prediction.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim depends on measured per-pulley efficiency values and on the transferability of those values to the robot. It uses standard linear algebra and quadratic programming assumptions. No new physical entities are introduced.

free parameters (3)

Per-pulley tension transmission efficiency eta_p = 0.919 to 0.979, tension dependent
Measured on the test stand at 200 N and 400 N with a 90 degree wrap for each wire-pulley pair; used in Eq. (5) through Eq. (7) to set the compensation level.
QP weighting matrix Lambda = not reported
Tuning parameter in the quadratic programs, Eq. (4) and Eq. (7); it changes the computed target wire tensions and thus the force error, but its values or selection rule are not given.
Tension bounds T_min and T_max = not reported
Constraints in the quadratic programs, Eq. (4) and Eq. (7); robot-specific values are not stated though they affect the tension allocation and the resulting force error.

assumptions (4)

standard math Standard vector calculus and Jacobian relationships, Eq. (1) through Eq. (3)
Used without proof to relate end-effector force, joint torque, and wire tension; not controversial.
domain assumption Muscle length Jacobian G(q) and joint Jacobian J_r are known exactly for the robot
The compensation accuracy depends on these kinematic models, and no identification error is reported in Section IV-C.
domain assumption Per-pulley efficiency is constant and multiplicative as eta_p^{N_ij} for every pulley on each wire path
Invoked in Eq. (5); assumes independence of tension, velocity, wrap angle, and direction, which is contradicted at reversals by cited work [16].
domain assumption Pre-stretching removes initial wire stretch sufficiently that residual creep is negligible during the force trials
Assumed in Section IV-C; no repeatability data after pre-stretch are shown.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Universal Wire Testing Machine for Enhancing the Performance of Wire-Driven Robots." pith.science (2026). https://pith.science/paper/5N22NNOB

@misc{pith2026250910862,
  author       = {Pith},
  title        = {Pith review of: A Universal Wire Testing Machine for Enhancing the Performance of Wire-Driven Robots},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5N22NNOB}},
  note         = {Machine review of arXiv:2509.10862}
}

read the original abstract

Compared with gears and linkages, wires constitute a lightweight, low-friction transmission mechanism. However, because wires are flexible materials, they tend to introduce large modeling errors, and their adoption in industrial and research robots remains limited.In this study, we built a Universal Wire Testing Machine that enables measurement and adjustment of wire characteristics to improve the performance of wire-driven mechanisms. Using this testing machine, we carried out removal of initial wire stretch, measurement of tension transmission efficiency for eight different diameters of passive pulleys, and measurement of the dynamic behavior of variable-length wires. Finally, we applied the data obtained from this testing machine to the force control of an actual wire-driven robot, reducing the end-effector force error.

Figures

Figures reproduced from arXiv: 2509.10862 by the authors.

**Figure 2.** Design overview of wire testing machine. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Detailed design of Multi-Pulley Pre-Stretching System. [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

Figures from the paper (6 more)

**Figure 4.** Figure 4: Snapshot of wire elongation test. The wire elongated from 8 [PITH_FULL_IMAGE:figures/full_fig_p003_4.png]

**Figure 5.** Figure 5: Detailed design of Passive Pulley Transmission Efficiency Mea [PITH_FULL_IMAGE:figures/full_fig_p004_5.png]

**Figure 9.** Figure 9: Frequency response of tension for constant and variable total wire [PITH_FULL_IMAGE:figures/full_fig_p005_9.png]

**Figure 8.** Figure 8: The relationship between pulley diameter and tension transmission [PITH_FULL_IMAGE:figures/full_fig_p005_8.png]

**Figure 10.** Figure 10: Setup of end-effector force measurement experiment. [PITH_FULL_IMAGE:figures/full_fig_p006_10.png]

**Figure 11.** Figure 11: Comparison of end-effector force error with and without tension [PITH_FULL_IMAGE:figures/full_fig_p006_11.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 18 canonical work pages

[16]

Tension analysis of a 6-degree- of-freedom cable-driven parallel robot considering dynamic pulley bearing friction,

S.-H. Choi, J.-O. Park, and K.-S. Park, “Tension analysis of a 6-degree- of-freedom cable-driven parallel robot considering dynamic pulley bearing friction,”Advances in Mechanical Engineering, vol. 9, no. 8, p. 1687814017714981, 2017

work page 2017
[1]

W AM Arm,

“W AM Arm,” https://web.barrett.com/files/W AMDataSheet 02.2011. pdf

work page 2011
[2]

Anthropomorphic low-inertia high-stiffness manipulator for high-speed safe interaction,

Y .-J. Kim, “Anthropomorphic low-inertia high-stiffness manipulator for high-speed safe interaction,”IEEE Transactions on robotics, vol. 33, no. 6, pp. 1358–1374, 2017

2017
[3]

Development of low-inertia high-stiffness manipulator lims2 for high- speed manipulation of foldable objects,

H. Song, Y .-S. Kim, J. Yoon, S.-H. Yun, J. Seo, and Y .-J. Kim, “Development of low-inertia high-stiffness manipulator lims2 for high- speed manipulation of foldable objects,” inProceedings of the 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2018, pp. 4145–4151

work page 2018
[4]

Coupled tendon-driven multijoint manipulator,

S. Hirose and S. Ma, “Coupled tendon-driven multijoint manipulator,” inProceedings of the 1991 IEEE International Conference on Robotics and Automation, 1991, pp. 1268–1275

work page 1991
[5]

Design and control of a seven-degrees-of-freedom manipulator actuated by a coupled tendon-driven system,

K. Yokoi, K. Tanie, N. Inamura, T. Kawai, and K. Agou, “Design and control of a seven-degrees-of-freedom manipulator actuated by a coupled tendon-driven system,” inProceedings IROS’91: IEEE/RSJ International Workshop on Intelligent Robots and Systems’ 91, 1991, pp. 737–742

work page 1991
[6]

Saqiel: Ultra-light and safe manipulator with passive 3d wire align- ment mechanism,

T. Suzuki, M. Bando, K. Kawaharazuka, K. Okada, and I. Masayuki, “Saqiel: Ultra-light and safe manipulator with passive 3d wire align- ment mechanism,”IEEE Robotics and Automation Letters, vol. 9, no. 4, pp. 1–8, 2024

work page 2024
[7]

Ramiel: A parallel-wire driven monopedal robot for high and con- tinuous jumping,

T. Suzuki, Y . Toshimitsu, Y . Nagamatsu, K. Kawaharazuka, A. Miki, Y . Ribayashi, M. Bando, K. Kojima, Y . Kakiuchi, K. Okada,et al., “Ramiel: A parallel-wire driven monopedal robot for high and con- tinuous jumping,” inProceedings of the 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 2022, pp. 5017– 5024

work page 2022

Show all 19 references

[8]

Overcoming physical limitations utilizing the surround- ing environment with a wire-driven multipurpose robot,

S. Inoue, K. Kawaharazuka, T. Suzuki, S. Yuzaki, K. Okada, and M. Inaba, “Overcoming physical limitations utilizing the surround- ing environment with a wire-driven multipurpose robot,”Advanced Robotics Research, vol. 1, no. 1, p. 2400021, 2025

2025
[9]

Design method of a kangaroo robot with high power legs and an articulated soft tail,

S. Yoshimura, T. Suzuki, M. Bando, S. Yuzaki, K. Kawaharazuka, K. Okada, and M. Inaba, “Design method of a kangaroo robot with high power legs and an articulated soft tail,” inProceedings of the 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2023, pp...

2023
[10]

Design optimization of wire arrangement with variable relay points in numerical simulation for tendon-driven robots,

K. Kawaharazuka, S. Yoshimura, T. Suzuki, K. Okada, and M. Inaba, “Design optimization of wire arrangement with variable relay points in numerical simulation for tendon-driven robots,”IEEE Robotics and Automation Letters, vol. 9, no. 2, pp. 1388–1395, 2023

2023
[11]

Measure- ment of the cable-pulley coulomb and viscous friction for a cable- driven surgical robotic system,

M. Miyasaka, J. Matheson, A. Lewis, and B. Hannaford, “Measure- ment of the cable-pulley coulomb and viscous friction for a cable- driven surgical robotic system,” inProceedings of the 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 2015, pp. 804–810

2015
[12]

Measurement of the tension loss in a ca- ble traveling over a pulley, for low-speed applications,

P. M ´at´e and A. Szekr ´enyes, “Measurement of the tension loss in a ca- ble traveling over a pulley, for low-speed applications,”Experimental Mechanics, vol. 64, no. 9, pp. 1579–1599, 2024

2024
[13]

Modeling of synthetic fiber ropes and frequency response of long-distance cable–pulley system,

A. Takata, G. Endo, K. Suzumori, H. Nabae, Y . Mizutani, and Y . Suzuki, “Modeling of synthetic fiber ropes and frequency response of long-distance cable–pulley system,”IEEE Robotics and Automation Letters, vol. 3, no. 3, pp. 1743–1750, 2018

2018
[14]

Super Dragon: A 10-m-Long- Coupled Tendon-Driven Articulated Manipulator,

G. Endo, A. Horigome, and A. Takata, “Super Dragon: A 10-m-Long- Coupled Tendon-Driven Articulated Manipulator,”IEEE Robotics and Automation Letters, vol. 4, no. 2, pp. 934–941, 2019

2019
[15]

Cable behavior influence on cable-driven parallel robots vibrations: experimental characterization and simulation,

D. Gueners, B.-C. Bouzgarrou, and H. Chanal, “Cable behavior influence on cable-driven parallel robots vibrations: experimental characterization and simulation,”Journal of Mechanisms and Robotics, vol. 13, no. 4, p. 041003, 2021

2021
[17]

A solid friction model,

P. DAHL, “A solid friction model,”Rep.TOR-0158(3107-18)-1, 1968

1968
[18]

Bearing Frictional coefficient,

“Bearing Frictional coefficient,” https://koyo.jtekt.co.jp/en/support/ bearing-knowledge/8-4000.html
[19]

A joint-space controller based on redundant muscle tension for multiple dof joints in musculoskeletal humanoids,

M. Kawamura, S. Ookubo, Y . Asano, T. Kozuki, K. Okada, and M. Inaba, “A joint-space controller based on redundant muscle tension for multiple dof joints in musculoskeletal humanoids,” inProceedings of the 2016 IEEE-RAS International Conference on Humanoid Robots, 2016, pp. 814–819

2016

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [16]

Tension analysis of a 6-degree- of-freedom cable-driven parallel robot considering dynamic pulley bearing friction,

S.-H. Choi, J.-O. Park, and K.-S. Park, “Tension analysis of a 6-degree- of-freedom cable-driven parallel robot considering dynamic pulley bearing friction,”Advances in Mechanical Engineering, vol. 9, no. 8, p. 1687814017714981, 2017

work page 2017

[2] [1]

W AM Arm,

“W AM Arm,” https://web.barrett.com/files/W AMDataSheet 02.2011. pdf

work page 2011

[3] [2]

Anthropomorphic low-inertia high-stiffness manipulator for high-speed safe interaction,

Y .-J. Kim, “Anthropomorphic low-inertia high-stiffness manipulator for high-speed safe interaction,”IEEE Transactions on robotics, vol. 33, no. 6, pp. 1358–1374, 2017

2017

[4] [3]

Development of low-inertia high-stiffness manipulator lims2 for high- speed manipulation of foldable objects,

H. Song, Y .-S. Kim, J. Yoon, S.-H. Yun, J. Seo, and Y .-J. Kim, “Development of low-inertia high-stiffness manipulator lims2 for high- speed manipulation of foldable objects,” inProceedings of the 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2018, pp. 4145–4151

work page 2018

[5] [4]

Coupled tendon-driven multijoint manipulator,

S. Hirose and S. Ma, “Coupled tendon-driven multijoint manipulator,” inProceedings of the 1991 IEEE International Conference on Robotics and Automation, 1991, pp. 1268–1275

work page 1991

[6] [5]

Design and control of a seven-degrees-of-freedom manipulator actuated by a coupled tendon-driven system,

K. Yokoi, K. Tanie, N. Inamura, T. Kawai, and K. Agou, “Design and control of a seven-degrees-of-freedom manipulator actuated by a coupled tendon-driven system,” inProceedings IROS’91: IEEE/RSJ International Workshop on Intelligent Robots and Systems’ 91, 1991, pp. 737–742

work page 1991

[7] [6]

Saqiel: Ultra-light and safe manipulator with passive 3d wire align- ment mechanism,

T. Suzuki, M. Bando, K. Kawaharazuka, K. Okada, and I. Masayuki, “Saqiel: Ultra-light and safe manipulator with passive 3d wire align- ment mechanism,”IEEE Robotics and Automation Letters, vol. 9, no. 4, pp. 1–8, 2024

work page 2024

[8] [7]

Ramiel: A parallel-wire driven monopedal robot for high and con- tinuous jumping,

T. Suzuki, Y . Toshimitsu, Y . Nagamatsu, K. Kawaharazuka, A. Miki, Y . Ribayashi, M. Bando, K. Kojima, Y . Kakiuchi, K. Okada,et al., “Ramiel: A parallel-wire driven monopedal robot for high and con- tinuous jumping,” inProceedings of the 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 2022, pp. 5017– 5024

work page 2022