Switch-based Independent Antagonist Actuation with a Single Motor for a Soft Exosuit

Atharva Vadeyar; Dario Farina; Etienne Burdet; Rejin John Varghese

REVIEW 3 major objections 6 minor 10 references

Switch-based Independent Antagonist Actuation with a Single Motor for a Soft Exosuit

T0 review · 3 major / 6 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read A switch-based gear mechanism lets one motor independently actuate both cables of an antagonist pair, regardless of cable path geometry, with a measured cable-switching time of 298.24 ms.

desk verdict A simple, genuinely new floating-gear switch that lets one motor drive an antagonist cable pair, with a solid qualitative proof-of-concept but a switching-time number that measures less than it claims. read the letter →

arxiv 2502.05290 v1 pith:IJQBPGDP submitted 2025-02-07 cs.RO

Atharva Vadeyar , Rejin John Varghese , Etienne Burdet , Dario Farina This is my paper · ORCID

classification cs.RO

keywords softexosuitcable-drivenactuationantagonistswitch-basedmechanismsingle-motoractuatorunderactuationwearableroboticsrehabilitation

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

Soft exosuits often need two motors per joint because the agonist and antagonist cables follow different paths and therefore need independent displacement control. This paper proposes a single-motor alternative: a switch gear that stays meshed with the motor and rolls along a track to engage either of two spring-loaded spools, so one motor can pull one cable while the other unwinds freely. The authors built a 56 g prototype, showed it drives a one-degree-of-freedom joint through its full range with geometrically different cable paths, and measured a cable-switching time of about 300 ms at 120 rpm. The intended benefit is generality: the actuator does not require custom pulley profiles, so the same hardware could be reused across different joints and users. This matters for multi-joint exosuits, where reducing the number of motors is a key design goal.

What carries the argument

The switch gear is a planet gear that always remains meshed with the motor's driving gear but is free to roll along a circular track instead of being fixed in place. It travels between two driven gears, each connected to a cable spool; the driven gears are spring-loaded with clock springs so the disengaged cable remains taut. Tooth friction between the driving gear and the switch gear powers the traversal, and the angular separation θ between the driven gears, together with the gear radii, determines the switching delay, which the paper expresses as proportional to θ divided by motor speed. This gear-in-a-track mechanism is what converts one motor direction into either of two cable pulls without requiring any cable-path-specific geometry.

What would settle it

Time a full cable-to-cable switch on the same test rig by commanding a direction reversal while recording both string potentiometer signals, measuring from the last sample where the inactive cable's displacement stays constant to the first sample where the formerly inactive cable begins to move; if this interval is substantially longer than the ~300 ms traversal time, the paper's switching-time measurement would understate the true latency.

Watch

Extended reading notes

Core claim

The central discovery is that a truncated planetary gear arrangement can act as a switch between two antagonistic cables driven by one motor. When the motor reverses direction, the switch gear is carried around the driving gear by tooth friction until it meshes with the opposite driven gear, which is attached to the other cable's spool; because the driven gears are spring-loaded and mechanically independent, the cable that is not being pulled can unwind freely without affecting the driven cable. The authors verify this independence by driving the joint over its full −90° to 90° range while perturbing the unwinding spool and observing no effect on the driven cable. They measure a switch traversal time of 298–302 ms over ten trials for each direction at approximately 120 rpm motor output, and note that the motor shaft rotates 122.6° while the switch gear revolves only 19.8°, so the switching delay is set by gear geometry and motor speed rather than by the cable path shapes.

Load-bearing premise

The reported 298 ms switching time was measured by driving the switch gear along the track without engaging either driven gear or spool, so it counts only the gear's free traversal; if gear meshing and the beginning of spool rotation add meaningful time, the actual cable-switch latency would be larger than the headline number.

Editorial extensions

If this is right

One actuation unit can replace two motors for each one-degree-of-freedom joint, halving the actuator count needed per joint in multi-joint soft exosuits.
Because the spools are independently spring-loaded, the mechanism works without a known fixed relationship between agonist and antagonist cable paths, allowing the same hardware to accommodate different routing geometries.
The roughly 300 ms switching latency is compatible with the 1–3 Hz movement rates typical of rehabilitation and activities of daily living, although it is too slow for highly dynamic assistance tasks.
Moving the switch gear to the center of the track disengages the motor from both spools, providing a built-in safety feature that lets the joint move transparently with no actuation.
Switching latency can be reduced in future versions by choosing a faster motor or adjusting the gear ratio between the driving and switch gears, since the delay scales with θ over motor speed.
The paper states that the mechanism's switching time is independent of gravity and that the friction between the driving and switch gear is sufficient to engage either spool reliably, with very low variability across trials.

Reading between the lines

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

The 298 ms figure measures only the switch gear's free traversal along the track, because the protocol explicitly ran the gear between two positions without engaging either driven gear or spool; the full cable-to-cable switching time would also include gear meshing and the start of spool rotation, which could add a nontrivial amount of time.
The same single-motor time-sharing principle could be extended beyond two driven gears, allowing one motor to sequentially actuate several cables or even multiple joints when the switching latency is acceptable for the task.
The claim of generality across 'any cable path geometry' is supported experimentally by one pair of curved paths; extending this to arbitrary, user-specific routings still relies on the spring-loaded spools providing enough tension to prevent slack, which is plausible but not exhaustively demonstrated.
Since the switch gear's motion is driven by the balance between tooth friction and track friction, the mechanism's reliability may depend on maintaining that friction balance over time—wear, lubrication, or manufacturing variations could affect the switching behavior.

Share X Bluesky LinkedIn Reddit HN

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,

Worth a look if you care about single-motor actuation for soft exosuits. The core idea is a translating floating gear, always meshed with the motor, that travels along a track to engage one of two spring-loaded driven spools. That is a real addition to the cited alternatives — task-defined pulleys, one-to-many actuation, differential drives — and it is mechanically simple enough to matter. They built a 56 g prototype and show, with a joint going through ±90°, that disturbing the unwinding cable does not affect the driven one. The decoupling works; the qualitative traces in Fig. 4 are consistent with what the mechanism should do.

The main soft spot is exactly the one the stress-test note flags. The 298 ms figure comes from timing the switch gear moving between track extremes without engaging either driven gear (Section III-B2). That is not the cable-to-cable switching latency advertised in the abstract. Disengagement, meshing into the new driven gear, and loading the spring-loaded spool are all excluded. The sub-millisecond sigma only says the free traversal is repeatable. This is load-bearing because the paper uses 300 ms to argue that 1–3 Hz ADL assistance is unaffected. At 3 Hz, even the free-traversal time is longer than the half-period, so the full switching event would make the limitation worse, not better. That claim needs to be walked back or retested end-to-end.

Two smaller issues. The independence test is purely qualitative; a force/tension measurement or a crosstalk number would make it convincing. And "any cable path geometry" is too strong for one test rig with two curved paths and a single joint. The mechanism plausibly generalizes, but the paper has not shown that.

The literature handling is fine. The only self-citation is a general wearable-robotics review, and it is not doing analytical work. Limitations are honestly stated, including the lack of simultaneous torque and stiffness control.

Who is this for? Someone building multi-joint cable-driven exosuits and looking for a way to halve actuator count. As a concept-validation paper it deserves a serious referee; the mechanism is novel, the prototype exists, and the flaws are fixable with additional measurements. I would not desk-reject it, but I would send it back for a proper end-to-end switching-time test and a toned-down generalization claim.

Referee Report

3 major / 6 minor

Summary. This paper proposes a single-motor, switch-based actuation mechanism for antagonist cable pairs in soft exosuits. A motor-driven 'switch gear' travels along a track and alternately engages one of two spring-loaded driven gears, allowing one motor to independently wind and unwind two cables that follow different curved paths. The authors validate the concept with two experiments: a spool-independence test measuring joint motion under random perturbations, and a switching-time test reporting a mean traversal time of about 298 ms at 120 rpm. The paper concludes that the mechanism is lightweight, generalizable to arbitrary cable path geometries, and adequate for ADL and rehabilitation applications despite its switching latency.

Significance. If the mechanism performs as claimed, it offers a low-complexity, lightweight alternative to dual-motor actuation, custom pulley profiles, and one-to-many actuators for multi-DoF soft exosuits. The decoupled-spool design is an elegant idea, and the qualitative demonstration of agonist/antagonist independence is valuable. The paper also honestly acknowledges limitations such as the inability to provide on-demand stiffness. However, the quantitative contribution rests on a measured quantity that does not match the claimed quantity, and the ADL frequency statement is internally inconsistent. With proper end-to-end timing and tempered claims, the work would be a useful contribution.

major comments (3)

[III-B2 and III-C] The reported switching time of 298.24 ms is not a cable-to-cable switching latency. The protocol in Section III-B2 times only the switch gear's free traversal between track extremes 'without engaging either driven gear/spool.' This omits disengagement from the active driven gear, meshing with the target driven gear, and the transient during which the newly engaged spool accelerates against the clock-spring and cable tension. Because the abstract presents this number as the time 'needed for switching between cables,' the central quantitative claim is unsupported. The authors should measure the full event end-to-end, e.g., from motor command reversal to the first detectable displacement of the newly driven spool, and report that latency separately from the free-traversal time.
[IV and Abstract] The claim that a ~300 ms switching delay is acceptable for 1–3 Hz ADL is internally inconsistent. At 3 Hz the half-cycle duration is 167 ms, which is already shorter than the reported 298 ms free-traversal time, so alternating between agonist and antagonist at that rate is impossible. Even at 1 Hz, a 300 ms switch leaves only 200 ms of active actuation per half-cycle, which is likely insufficient for smooth assistance. The paper should either state the maximum switching frequency compatible with the measured delay, substantially reduce the latency, or qualify the ADL statement to very slow movements.
[Abstract and III-A] The claim of actuating 'along any cable path geometry' is supported only by a single pair of curved cable paths in the test rig and a qualitative independence observation. While the decoupled-spool design is conceptually geometry-agnostic, the validation does not demonstrate that the clock-spring preload maintains cable tension and avoids slack across other geometries, such as paths with large length differences or non-monotonic routing. The claim should be either tempered to 'the tested paths' or accompanied by a formal argument for why any geometry is accommodated, ideally with additional experiments varying path geometry.

minor comments (6)

[Abstract] The sentence 'it can reduced in the future' contains a grammatical error; it should read 'it can be reduced in the future.'
[II] The term 'truncated planetary gear system' is not standard; please clarify what is truncated and why the planetary analogy is useful for readers unfamiliar with the arrangement.
[III-B2] The 'empirically determined points' for the switching-time test are not reported as motor positions or angles; reporting the commanded displacement would aid reproducibility.
[III-C] The spool-independence results are described only qualitatively with reference to Fig. 4; quantitative metrics such as maximum perturbation, correlation between spool displacements, or root-mean-square deviation would strengthen the claim.
[III-C and Fig. 5] Figure 5 is mentioned in the text but the reader cannot verify the reported means and standard deviations from the figure; consider adding error bars or individual trial points.
[II and IV] The safety feature of disengaging the motor by moving the switch gear to the center is claimed but not experimentally validated; a brief test or at least a specification of the commanded center position would support this feature.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the central claims rest on direct hardware experiments, not on fitting, self-referential derivation, or load-bearing self-citation.

full rationale

The paper's central claims are (i) that a switch-based mechanism can independently actuate an antagonist pair of cables from a single motor, and (ii) that the measured switching time is about 298 ms. Both claims are supported by direct physical experiments described in Section III: the spool-independence test directly demonstrates decoupled agonist/antagonist motion, and the switching-time test directly reports traversal durations of 302 ms and 298 ms over ten trials. No parameter is fitted to a subset of data and then renamed as a prediction; the switching time is a measured value, not a derived quantity. The paper does not invoke any uniqueness theorem, and it does not justify its design by an ansatz smuggled in through citation. The only author self-citation is reference [2], a general review of wearable robotics used for background motivation; no load-bearing argument depends on it. The skeptic concern that the switching-time protocol measures only free switch-gear traversal without engaging the driven gears or spools is a measurement-validity/correctness risk, not a circularity: the reported number is not equivalent by construction to its inputs, and the paper does not hide the protocol. Therefore the circularity score is 0.

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

The central claims (independent antagonist actuation, 300 ms switching, generalization to arbitrary paths) depend on unquantified friction and spring behaviors, plus an extrapolation from a single test rig. No free parameters are fitted; the mechanism is a physical design rather than a model.

assumptions (3)

domain assumption The friction between the driving gear and the switch gear is sufficient and reliable enough to make the switch gear revolve around the motor and engage the correct driven gear at the end of the track.
Stated in Section II and inferred from the switching experiments, but not quantitatively modeled or tested under varying loads, speeds, and surface wear.
domain assumption The clock-springs on both spools keep the disengaged cable taut, preventing slack and uncontrolled unspooling.
Described as a design feature in Section II, but its tension range and behavior under fast or repeated switching are not characterized.
ad hoc to paper The two curved cable paths in the test rig are representative of 'any cable path geometry' encountered in soft exosuits.
The abstract claims generality to any cable path geometry, but only one test rig with two curved profiles was evaluated.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Switch-based Independent Antagonist Actuation with a Single Motor for a Soft Exosuit." pith.science (2026). https://pith.science/paper/IJQBPGDP

@misc{pith2026250205290,
  author       = {Pith},
  title        = {Pith review of: Switch-based Independent Antagonist Actuation with a Single Motor for a Soft Exosuit},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IJQBPGDP}},
  note         = {Machine review of arXiv:2502.05290}
}

read the original abstract

The use of a cable-driven soft exosuit poses challenges with regards to the mechanical design of the actuation system, particularly when used for actuation along multiple degrees of freedom (DoF). The simplest general solution requires the use of two actuators to be capable of inducing movement along one DoF. However, this solution is not practical for the development of multi-joint exosuits. Reducing the number of actuators is a critical need in multi-DoF exosuits. We propose a switch-based mechanism to control an antagonist pair of cables such that it can actuate along any cable path geometry. The results showed that 298.24ms was needed for switching between cables. While this latency is relatively large, it can reduced in the future by a better choice of the motor used for actuation.

Figures

Figures reproduced from arXiv: 2502.05290 by the authors.

**Figure 2.** Schematic of the switch-based actuation mechanism: A). CAD rendering of the mechanism with labelled components, B). Geometrical configuration [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 3.** The test rig to test the switching mechanism. It consists of a motor with the switching mechanism. The capstans in the mechanism are connected [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 5.** Figure 5: Switching Time Test Results [3] H. In, H. Lee, U. Jeong, B. B. Kang, and K.-J. Cho, “Feasibility study of a slack enabling actuator for actuating tendon-driven soft wearable robot without pretension,” in 2015 IEEE International Conference on Robotics and Automation (IC…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

10 extracted references · 10 canonical work pages

[1]

11em plus .33em minus .07em @technote 4000 4000 100 4000 4000 500 `\.=1000 = #1 #1 #1 0pt [0pt][0pt] #1 * \| ** #1 \@IEEEauthorblockNstyle \@IEEEauthorblockAstyle \@IEEEauthordefaulttextstyle \@IEEEauthorblockconfadjspace -0.25em \@IEEEauthorblockNtopspace 0.0ex \@IEEEauthorblockAtopspace 0.0ex \@IEEEauthorblockNinterlinespace 2.6ex \@IEEEauthorblockAinte...

work page
[2]

Xiloyannis, R

M. Xiloyannis, R. Alicea, A.-M. Georgarakis, F. L. Haufe, P. Wolf, L. Masia, and R. Riener, ``Soft robotic suits: State of the art, core technologies, and open challenges,'' IEEE Transactions on Robotics, vol. 38, no. 3, pp. 1343--1362, 2021

work page 2021
[3]

R. J. Varghese, D. Freer, F. Deligianni, J. Liu, G.-Z. Yang, and R. Tong, ``Wearable robotics for upper-limb rehabilitation and assistance: A review of the state-of-the-art challenges and future research,'' Wearable technology in medicine and health care, pp. 23--69, 2018

work page 2018
[4]

H. In, H. Lee, U. Jeong, B. B. Kang, and K.-J. Cho, ``Feasibility study of a slack enabling actuator for actuating tendon-driven soft wearable robot without pretension,'' in 2015 IEEE International Conference on Robotics and Automation (ICRA). 1em plus 0.5em minus 0.4em IEEE, 2015, pp. 1229--1234

work page 2015
[5]

H. In, U. Jeong, H. Lee, and K.-J. Cho, ``A novel slack-enabling tendon drive that improves efficiency, size, and safety in soft wearable robots,'' IEEE/ASME Transactions on Mechatronics, vol. 22, no. 1, pp. 59--70, 2016

work page 2016
[6]

Xiloyannis, L

M. Xiloyannis, L. Cappello, D. B. Khanh, S.-C. Yen, and L. Masia, ``Modelling and design of a synergy-based actuator for a tendon-driven soft robotic glove,'' in 2016 6th IEEE International Conference on Biomedical Robotics and Biomechatronics (BioRob). 1em plus 0.5em minus 0.4em IEEE, 2016, pp. 1213--1219

work page 2016
[7]

Alicea, M

R. Alicea, M. Xiloyannis, D. Chiaradia, M. Barsotti, A. Frisoli, and L. Masia, ``A soft, synergy-based robotic glove for grasping assistance,'' Wearable Technologies, vol. 2, p. e4, 2021

work page 2021
[8]

Xiloyannis, E

M. Xiloyannis, E. Annese, M. Canesi, A. Kodiyan, A. Bicchi, S. Micera, A. Ajoudani, and L. Masia, ``Design and validation of a modular one-to-many actuator for a soft wearable exosuit,'' Frontiers in neurorobotics, vol. 13, p. 39, 2019

work page 2019

Show all 10 references

[9]

Bajaj, V

A. Bajaj, V. Jain, P. Kumar, A. Unal, and A. Saxena, ``Soft hand exoskeleton for adaptive grasping using a compact differential mechanism,'' in Mechanism and Machine Science: Select Proceedings of Asian MMS 2018. 1em plus 0.5em minus 0.4em Springer, 2020, pp. 733--746

2018
[10]

Zhang, B

W. Zhang, B. G. Cangan, T. Buchner, A. M. Kübler, R. Asmus, and R. K. Katzschmann, ``Task-defined pulley design for nonlinearly coupled tendon-driven actuation,'' in 2024 IEEE 7th International Conference on Soft Robotics (RoboSoft), 2024, pp. 220--227

2024

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.

[1] [1]

11em plus .33em minus .07em @technote 4000 4000 100 4000 4000 500 `\.=1000 = #1 #1 #1 0pt [0pt][0pt] #1 * \| ** #1 \@IEEEauthorblockNstyle \@IEEEauthorblockAstyle \@IEEEauthordefaulttextstyle \@IEEEauthorblockconfadjspace -0.25em \@IEEEauthorblockNtopspace 0.0ex \@IEEEauthorblockAtopspace 0.0ex \@IEEEauthorblockNinterlinespace 2.6ex \@IEEEauthorblockAinte...

work page

[2] [2]

Xiloyannis, R

M. Xiloyannis, R. Alicea, A.-M. Georgarakis, F. L. Haufe, P. Wolf, L. Masia, and R. Riener, ``Soft robotic suits: State of the art, core technologies, and open challenges,'' IEEE Transactions on Robotics, vol. 38, no. 3, pp. 1343--1362, 2021

work page 2021

[3] [3]

R. J. Varghese, D. Freer, F. Deligianni, J. Liu, G.-Z. Yang, and R. Tong, ``Wearable robotics for upper-limb rehabilitation and assistance: A review of the state-of-the-art challenges and future research,'' Wearable technology in medicine and health care, pp. 23--69, 2018

work page 2018

[4] [4]

H. In, H. Lee, U. Jeong, B. B. Kang, and K.-J. Cho, ``Feasibility study of a slack enabling actuator for actuating tendon-driven soft wearable robot without pretension,'' in 2015 IEEE International Conference on Robotics and Automation (ICRA). 1em plus 0.5em minus 0.4em IEEE, 2015, pp. 1229--1234

work page 2015

[5] [5]

H. In, U. Jeong, H. Lee, and K.-J. Cho, ``A novel slack-enabling tendon drive that improves efficiency, size, and safety in soft wearable robots,'' IEEE/ASME Transactions on Mechatronics, vol. 22, no. 1, pp. 59--70, 2016

work page 2016

[6] [6]

Xiloyannis, L

M. Xiloyannis, L. Cappello, D. B. Khanh, S.-C. Yen, and L. Masia, ``Modelling and design of a synergy-based actuator for a tendon-driven soft robotic glove,'' in 2016 6th IEEE International Conference on Biomedical Robotics and Biomechatronics (BioRob). 1em plus 0.5em minus 0.4em IEEE, 2016, pp. 1213--1219

work page 2016

[7] [7]

Alicea, M

R. Alicea, M. Xiloyannis, D. Chiaradia, M. Barsotti, A. Frisoli, and L. Masia, ``A soft, synergy-based robotic glove for grasping assistance,'' Wearable Technologies, vol. 2, p. e4, 2021

work page 2021

[8] [8]

Xiloyannis, E

M. Xiloyannis, E. Annese, M. Canesi, A. Kodiyan, A. Bicchi, S. Micera, A. Ajoudani, and L. Masia, ``Design and validation of a modular one-to-many actuator for a soft wearable exosuit,'' Frontiers in neurorobotics, vol. 13, p. 39, 2019

work page 2019