On the Design of a Simulation-Assisted Human-Centered Quasi-Stiffness-Based Actuator for Ankle Orthosis

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript details the development and results of an open-source simulation method for designing ankle exoskeleton assistance during the stance phase. Specifically it leverages prior work within Open Sim and MatLab toolboxes from the biomechanics literature to predict spring stiffness and torque outputs for elastic actuator designs of ankle exoskeletons during loaded torso (backpack) walking for flat ground based on joint quasi-stiffness assistance. Given the gap created by the recent developments within Open Sim that have made prior prior models in this area unusable, this work provides a much needed solution to restore this design capability. The detail provided within the manuscript is outstanding but could still benefit from making a few points clearer or stating the limitations better. 

Comments:

1) Please make clear if the mass of the actuator is being included in the simulation and if so, at what location on the body. If it is being included, state what motors are representative of delivering those torques with those device weights so that the results may seem plausible. If the actuator mass is not being modeled, then state that explicitly in the conclusions.

2) The impact of magnitude and locations of added mass on the body (backpack, orthosis, and actuator) with respect to gait adaptations needs to be addressed more clearly. Specifically, is the simulation yielding a change in the gait when the backpack (or other passive objects) are worn? If so, how was this determined and if not, can the errors be estimated so that the simulations are still deemed accurate? Some citations to consider for the impact of added mass on gait are the following: 

Spatiotemporal and muscle activation adaptations during overground walking in response to lower body added mass V Vijayan, S Fang, T Reissman, AL Kinney… - Gait & Posture, 2022  

3) Changing the stiffness at the ankle orthosis will influence other joint moments. Please clarify how these results compare with the literature. A reference is the following: 

T. Kobayashi, M.L. Singer, M.S. Orendurff, F. Gao, W.K. Daly, K.B. Foreman The effect of changing plantarflexion resistive moment of an articulated ankle-foot orthosis on ankle and knee joint angles and moments while walking in patients post stroke Clin. Biomech., 30 (2015), pp. 775-780, 10.1016/j.clinbiomech.2015.06.014  

4) Suggest a minor change to the title, as the design is not directly investigating variable stiffness but instead tuning to particular individuals fixed stiffness designs that are based on quasi-stiffness estimates. Thus suggest removing the words "variable stiffness" and instead insert at same location "quasi-stiffness based"  

Author Response

Dear reviewer, 

I am most grateful for the time you have taken to review our paper.

Please find enclosed a PDF file containing the responses to your comments. This document responds to the comments of both Reviewer 1 and Reviewer 2, as the authors believe that the reviewers may be interested in each other's comments and responses.

Yours sincerely.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript presents a comprehensive approach to designing an ankle orthosis using simulations and human-centered design principles. However, it is necessary to clarify the technical workflow (especially in sections like the biomechanical model setup) might improve readability for broader audiences, such as clinical practitioners or interdisciplinary researchers. Also, the authors are suggested to incorporate the following comments and suggestions for improving the manuscript.

1.        The methodology is well-documented, particularly the use of OpenSim and Matlab/Simulink. However, there is some lack of clarity in the transition between the initial simulation results and the practical prototyping suggestions. More details on the proposed mechanical components used in simulations (e.g., the specific spring and motor specifications) would help substantiate the claims.

2.        It will be great to add further explanation regarding Residual Reduction Algorithm (RRA) and how it specifically benefits simulation results in the presented work. A concise explanation may aid readers unfamiliar with the touches of kinematic adjustments in simulations.

3.        While the manuscript addresses current challenges in exoskeleton and orthosis design, it would be helpful to elaborate more on how the proposed method compares with existing orthoses (such as HULC or MindWalker). How does the proposed solution improve upon them, particularly in terms of energy efficiency, weight reduction, and wearability?

4.        The paper highlights inter-subject variations in ankle quasi-stiffness (AQS) profiles. Could this variability require different designs for each user, or is there potential for a more universal design that could be adjusted in real-time? How does the simulation account for these differences?

5.        The manuscript acknowledges that the ground reaction force and moment (GRFM) estimation using the CusToM toolbox may introduce discrepancies, particularly due to parameter mismatches between models. The authors suggested including a clarification of how significant these discrepancies are, and their impact on the overall conclusions.

6.        The authors mention a threshold of 5% error for GRFM estimation. What specific effects do these errors have on the simulated assistance profiles, and how might these deviations manifest in a physical system?

7.        The RMS fitting error presented is useful, but how does this compare with other simulation or hardware prototypes in the literature? For instance, how significant is the 0.0377 Nm/kg error when considering real-world usage scenarios?

8.        The simulation integrates active and passive actuators. In practice, how does the system decide when to prioritize one over the other? Are there any testing scenarios where the active component (torque generator) fails or is undersupplied?

9.        The manuscript provides a link to the open-source simulation tool. What kind of user base is this aimed at? Do the authors anticipate non-technical users (e.g., clinicians) will be able to utilize the tool effectively, or would it require a specific set of technical skills?

10.   Finally, given the promising simulation results, what are the anticipated challenges in miniaturizing the components for real-world usage (e.g., battery life, actuator power, or orthosis wearability)? How will this impact the portability and comfort of the device for extended use?

Author Response

Dear reviewer, 

I am most grateful for the time you have taken to review our paper.

Please find enclosed a PDF file containing the responses to your comments. This document responds to the comments of both Reviewer 1 and Reviewer 2, as the authors believe that the reviewers may be interested in each other's comments and responses.

Yours sincerely.

Author Response File: Author Response.pdf