A Novel Principle for Transparent Applications of Force Impulses in Cable-Driven Rehabilitation Systems

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This work proposed a new cable-driven actuation mechanism designed for
rehabilitation devices. Its main novelty lies in the fact that it decouples the
motion of the cable from the actuator via a reel-swivel design, paired with a
tension spring, resulting in a mechanism that allows the cable to freely moving
when assistive force is not needed, and thereby more "transparent" to the user.
Both bench tests and walking trial with human subjects were conducted, the
result suggested that the device can provide proper assistance at the push-off
phase, while remaining largely transparent when assistance is not needed.

The manuscript is well organized and easy to follow. The proposed design is
unique, and has great potential to benefit rehabilitation robotics in general.
However, the reviewer believes there are a few nontrivial flaws in the design of
the device, and the design of the experiments, that weakened impactfulness of
the work. Please see more comments below.

Major comments:
1) Uncompensated tension spring forces during the "idle" phase
Based on basic Hooke's law, the force applied by the tension spring increases
as the displacement increases - from the reviewer's understanding, the proposed
device does not actively compensate for such changes. As a result, the wearer
will feel increased force during the gait phases where "extended" cable lengths
are needed. In this case, ankle dorsiflection during the swing phase. The
reviewer believes the analysis in this critical aspect is overlooked.
As a reference, a very similar idea was proposed and validated in [1], where the
"semi-active" mode uses a spring to handle the "slack" cable, while the varying
spring forces resulted from changing displacement were constantly monitored and
compensated by the actuator.

2) Analysis of "transparency" is conducted in an inappropriate manner
According to L351-354, the analysis of the device's transparency was conducted
at gait phases "outside" of the push-off assistance - this is the opposite of
what the assist-as-needed paradigm aims to achieve. By design, the device should
remain as transparent as possible _during_ the phases where the patient need the
assistance the most, hence assist "as needed", instead of the phases where
patients do not necessarily need any assistance, if at all.

3) More detailed analysis of the device's limitation is needed
From the reviewer's understanding, the rate at which the cable can
unwind/retract during the M_ACTIVE -> M_PASSIVE phase is a critical factor that
affects whether the mechanism impedes the wearer's natural motion - if the
wearer's current motion necessitates rapid unwinding at a rate that's faster
than the motor's ability to back-drive, the wearer's motion will be effectively
"blocked" in an abrupt manner. Analysis on the maximum pace at which the
mechanism can safely support the intended motion needs to be investigated.
Note that this factor cannot be captured by the current bench testing setup
explained in Section 2.4 due to the static nature of the cable mounting point at
the distal end. The t_return result listed in L386 partially covered this
behavior, yet only from a temporal point of view.

4) Difference in key element used in bench testing v.s. human walking trials
An "elastic element" was introduced for the latter, at L252-254, yet none of its
characteristics was analyzed, which renders the bench test result not as
convincing and representative.

Minor comments:
- consider adding pointed arrows to help illustrate components listed in Fig. 1)
- consider indexing the M_ACTIVE diagrams shown in Fig. 2a), and add more
  references such as alpha_FORK

[1] Y. Zhang, R. J. Kleinmann, K. J. Nolan and D. Zanotto, "Preliminary Validation of a Cable-Driven Powered Ankle–Foot Orthosis With Dual Actuation Mode," in IEEE Transactions on Medical Robotics and Bionics, vol. 1, no. 1, pp. 30-37, Feb. 2019

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This study described development and evaluation of a PACE-R (Passive Active Cable Robot) for ankle assistance. PACE-R includes a spring, a fabric-based cable, a motor, a fork reel, a fork swievel and a drum. The gap between the fork swievel and reel enables coupling and decoupling of the motor power and the cable force. Application of PACE-R was demostrated in a treadmill test, which involved an additional elastic element (looks like a spring), and a Bowden cable connecting to an ankle orthosis. It was said that the overall system produced controlled assistance in active mode and almost transparency in the passive mode during the ankle  push-off test on the treadmill. This work is interesting, but several questions should be clarified to improve the overall work:

1. The innovation of the work is the mechanical design of the system. One of the biggest limitations, as described in the paper, is that only open-loop torque control was implemented in the application. The application of the PACE-R used a spring with Bowden cable. Two springs worked together, then the force and position should be complicated within one gait cycle and especially during walking at different speeds. Then an impedance/admittance control is required to secure normal functionality. It would be good for the authors to highlight and emphasise the mechanical development, while a section for “Limitations” should be specifically presented regarding the control part.
2. Mechanical design: The structure of PACE-R is very interesting. But the special terms such as “fork swivel”, “fork reel” make it difficult to understand the working principle. What is not clear for me is whether the swivel would rotate upon its fixing axis so as to brake the cable during the active mode. If yes, then the force on the load side during active mode is clear: directly influenced by the torque output from the motor. But from Figures 1-2, it seems that the swivel is fixed. Then in this case, the force output should be related to the motor torque and the torsion spring. Therefore, an open-loop torque control is too weak to secure a proper function, because the actual force is dependent on the stiffness of the spring, the position of the cable, and also the friction between the fork swivel and cables. If there are several revolutions of the fork in the active mode, then the friction between different layers of cable is more complex. The study presented the test on the Benchtop, which showed good relationship between force and torque. But as the author said, in the real application, “such conditions are unlikely in rehabilitation setting”. In the physical application, a second spring with a Bowden cable is introduced, then the dynamics yielded from the Benchtop experiment doesn’t apply anymore. Therefore, a closed-loop force or position control is required. Maybe a video for the evaluation test on the treadmill is helpful to help the readers to understand the working principle of PACE-R.
3. Suggestion: as the highlight of this work, the mechanical part should be described clearer, such as the diameter of the fork reel, fixation of the swivel, determination of the torsion spring, the maximal force, speed or position output of PACE-R in active mode, usage constraints, such as whether it is only used within one revolution of the reel or not.
4. The ankle orthosis: During the power calcuation, is the weight of the orthosis considered? If it is rigid, the relative movement between the shoe and the orthsis is very common. How was the the orthosis fixed on the shoe of the participant. The power calculation Eq. 5: during walking, the effective level arm r_wire is changing with the actual ankle joint angle. This should also be considered in the calculation.
5. The treadmill test: The actual force information is missing. The actual froce on the participant is influenced by the force from the two springs, the motor and the ankle joint. Then the return time should be affected. How long did one gait cycle last?
6. The work only analysed the active phase and passive phase. What about the transparency during the transition phase?

Some minor points regarding the presentation:

Abstract section: this study described two tests: benchtop experiment and a treadmill push-off demonstration. Lots of results were presented in the result section. However, in the abstract, only few results are presented. Abstract should summarise the results presented in the paper.

Figs 1-4: The components should be marked and labeled normally as general figure instead of using different colors for different compoenents, and repeating the label number repetitively in the text.

The legends of Figs 8-10 are out of the figures. Furthermore, it is not appropriate to present the legend just in one subfigure. The legend of Fig. 11 is missing.

Comments on the Quality of English Language

English can be improved by removing the grammar mistakes.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

the design of the ankle joint exoskeleton looks interesting by supporting multiple modes of operation.  The combination of the fork and spring essentially composes a variable stiffness mechanism (VSMs). the authors are invited to investigate on the stiffness changes and discuss on the differences of their design with other VSMs.

the authors defined transparency in the paper, but references are missing. moreover, it is recommended to include their control law for transparency regulation.  

 

it is not clear how many subjects were involved in the experiments. this information has to be clearly stated. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

This paper introduces the so-called PACE-R module, an actuation system designed to deliver targeted force impulses to human body segments. The cable-driven mechanism employs remote actuation with an innovation: mechanical coupling between the cable and motor that occurs only during active force application. For passive operation, the cable remains decoupled from the motor, with minimal tension maintained via a low-stiffness spring. Authors claim this design eliminates the detrimental effects of inertia, friction, and control-related latencies associated with motors, mechanical components, and sensory systems. However, these claims cannot be confirmed in the conclusion section, nor is it evident how much these characteristics have been improved. Moreover, my main concern is the delivered introduction: the main contribution of the paper related to the existing literature is not clear. What advantages does the proposed system offer compared to those form the state-of-the-art? In my opinion, the contribution is not well-defined and the introduction should be substantially improved. Additionally, the references lack consistent formatting and could also benefit from updates.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

(null)

Author Response

There were no new comments from the reviewer.

We thank the reviewer for their constructive comments during the review process of this manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

The comments were addressed.

Author Response

Reviewer’s comment:

The comments were addressed.

Author’s response

We thank the reviewer for their constructive comments during the review process of this manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

the reviewer has no more comments. On the other hand, it would be desirable if more subjects could be involved. 

Author Response

Reviewer’s comment:

the reviewer has no more comments. On the other hand, it would be desirable if more subjects could be involved. 

Author’s response:

We thank the reviewer for their constructive comments during the review process of this manuscript.

Reviewer 4 Report

Comments and Suggestions for Authors

In [36], in which the correct year of publication is 2024, the authors extensively reviewed the literature regarding the methods for mimicking the anisometric contractions of human muscles and achieving system transparency, in terms of mechanical transmission and functionality, for the actuation system of wearable robotic devices. These references can be used to contrast the advantages of this proposal, and then the contribution could be well defined contrasting the literature and the introduction rewritten and improved.  

There are some typos in the document, it could be revised.

[36] M. Xu, Z. Zhou, Z. Wang, L. Ruan, J. Mai and Q. Wang, "Bioinspired Cable-Driven Actuation System for Wearable Robotic Devices: Design, Control, and Characterization," in IEEE Transactions on Robotics, vol. 40, pp. 520-539, 2024, doi: 10.1109/TRO.2023.3324200.

Author Response

We thank the reviewer for constructive comments, which we believe have significantly enhanced the overall quality of the manuscript. In the revised version, text highlighted in red corresponds to the first revision, while text highlighted in blue corresponds to the second revision.

 

Reviewer’s comment 1:

In [36], in which the correct year of publication is 2024, the authors extensively reviewed the literature regarding the methods for mimicking the anisometric contractions of human muscles and achieving system transparency, in terms of mechanical transmission and functionality, for the actuation system of wearable robotic devices. These references can be used to contrast the advantages of this proposal, and then the contribution could be well defined contrasting the literature and the introduction rewritten and improved.  

[36] M. Xu, Z. Zhou, Z. Wang, L. Ruan, J. Mai and Q. Wang, "Bioinspired Cable-Driven Actuation System for Wearable Robotic Devices: Design, Control, and Characterization," in IEEE Transactions on Robotics, vol. 40, pp. 520-539, 2024, doi: 10.1109/TRO.2023.3324200.

 

Author’s response 1:

Regarding the publication year: we used Zotero bibliography management system and for some reason the publication year was not correct. This was now corrected in the final version of the manuscript.

Regarding the anisometric contractions: as suggested by the reviewer, an overview of anisometric contractions with corresponding references has been added to the Introduction section to contrast the advantages of our system (Lines 85–98, in blue).

 

Reviewer’s comment 2:

There are some typos in the document, it could be revised.

 

Author’s response 2:

The manuscript was revised by a native English-speaking person.