Mechanical and Control Design of an Industrial Exoskeleton for Advanced Human Empowering in Heavy Parts Manipulation Tasks

Round  1

Reviewer 1 Report

The authors present the early stage development of an industrial exoskeleton covering the design of the mechanical structure and the control algorithms. 

On the overall I enjoyed reading the paper which to my knowledge seems complete. 

There are a few English language mistakes but with a careful re-reading of the paper they can be easily eliminated. 

Some minor suggestions: in figure 1 there is an ambiguity between configuration and position (when referring to the two main positions). they are defined as configurations, described in figure as positions and later referred either as positions or configurations. 

Also, it might be interesting and easier to understand for the reader if the authors could add a schematic representation of the 9 phases (0 - 8) of the task. 

Throughout the paper some of the sizes/weights of parts are mentioned but (if I didn't miss it) I couldn't find the overall weight of the system which would be I believe of interest as it would have to be carried as a backpack by the user. 

Also in the link length design an average human anthropometric characteristics are used. Have you considered a maximum value for the adjustment elements to fit a certain population? 

The mathematical formulations seem to be correct and they are validated by numerical simulations. 

As a comment, the paper seems complete to me also without section 6 which introduces the new concept with additional DoFs. As in the first paragraph it is specified that the new design intends to improve the shoulder mobility one might expect that the initial design has some limitations (thus the need of the new model). So this should be written in a way not to "cast a shadow" over the design which would be also developed into an experimental model. 

On the overall, I will say it again, I enjoyed the paper and congratulations to the authors for their work. 

Author Response

Dear reviewer,

authors would like to thank you for your expertise and careful revision.

Please find in attached the response to your questions/comments.

Authors hope to satisfy your requests with the provided replies and revision.

Author Response File: Author Response.pdf

Reviewer 2 Report

The low-cost exoskeleton, control design, and simulation are worthwhile contributions. The manuscript is technically sound and well prepared.

        I did notice one minor typographical error:

2.2.2. Torques Requirements - Torque should be singular in the section title and first sentence. 

Author Response

Dear reviewer,

authors would like to thank you for your expertise and careful revision.

Please find in attached the response to your questions/comments.

Authors hope to satisfy your requests with the provided replies and revision.

Author Response File: Author Response.pdf

Reviewer 3 Report

The contribution deals with the mechanical and control design of a simplified exoskeleton for heavy parts manipulation tasks. Accordingly., the introduction presents a short overview of existing exoskeletons and assigned control solutions, which should possibly be extended and generalized to scientific and industrial application. After introducing the exemplary manipulation task in section 2, the modelling and design of the exoskeleton is presented within section 3, summarizing the kinematics and dynamics model as well as the realized design approach. Subsequently, the control scheme is discussed and validated within sections 4 and 5. After introducing possible extensions of the shoulder joint (section 6), the contribution ends with the conclusions in section 7.

The used methodology and algorithms seem to be well-considered. Hence, the authors first introduce the robot kinematics and dynamics. Furthermore, the control scheme is presented based on optimal control and fuzzy logic. After all, the proposed design is tested and verified by simulation on a 2-DOF serial manipulator. Despite the comprehensible methodology, the suitability of the proposed solution should be questioned and confirmed by practical tests. It is worth considering the inclusion of the resulting prototype in this contribution, as this would support the feasibility of the assigned application and drastically improve the scientific content.

The language of the paper includes several mistakes in expression, grammar and punctuation. Furthermore, the contents, in some passages, is not expressed clearly. In order to provide a well-readable paper, the text should comprehensively be corrected within a major revision.

Author Response

Dear reviewer,

authors would like to thank you for your expertise and careful revision.

Please find in attached the response to your questions/comments.

Authors hope to satisfy your requests with the provided replies and revision.

Author Response File: Author Response.pdf

Reviewer 4 Report

This paper introduces a novel industrial exoskeleton for assisting heavy part manipulation tasks. The robotic design as well as the enabling algorithm “fuzzy logic controller” was introduced. Although the idea, the uniqueness and the difference compared to previously proposed similar systems are not discussed. The evaluation should be performed using the actual system in addition to the simulation study. These significantly limits the novelty and innovation of the work.

- Scientific contribution is not well defined. The described 2DOF configuration and calculation of dynamics sounds a typical configuration for such application.

- There are design components in this paper, but the novelty is not well emphasized compared to previously proposed powered suit products.

- Literature review should be thoroughly performed and compare the hardware and algorithmic improvement.

- This paper discusses different DOF separately between sections. If ultimate configuration is based on 4DOF, why don’t authors use 4DOF consistently throughout the paper?

- Actual user study must be performed to validate the system evaluation.

- Simulation setting is not well described. For Figure 12, what gain parameters and equations were used for PID control? What was the justification of choosing 10kg as the reference and why 8, 12 16Nm was chosen?

- Discussion section is missing.

Author Response

Dear reviewer,

authors would like to thank you for your expertise and careful revision.

Please find in attached the response to your questions/comments.

Authors hope to satisfy your requests with the provided replies and revision.

Author Response File: Author Response.pdf

Reviewer 5 Report

The authors present the mechanical  and control architecture overall design of an industrial exoskeleton thought to assist workers during specific manipulations tasks.

The authors proposed an interesting final point of arrival (the design and control of the whole exoskeleton) and the paper deals with a worthy point, however, in my opinion, it needs some changes to result suitable for publication.

GENERAL REMARKS

I strongly suggest stressing the novelty of the proposed solution in Section 1.3. Especially comparing the given design (both mechanical con control) with respect to the already available systems.

I have some concerns regarding the way the measurement units are defined throughout the text: sometimes they are given as [kg] (just to mention an example), sometimes without brackets. Please uniform the style. Personally, I would suggest removing the brackets according to the SI.

SPECIFIC COMMENTS
In section 1.1, the industrial exoskeletons are classified into two main classes: active and passive solutions. Active systems are given, passive solutions are instead almost missing, since only one reference (i.e., [4]) is provided. I suggest the author to improve the state of the art analysis.

In addition, because of the target of the proposed solution (industrial scenario), the assessment of the market (or the research state of the art) results really important to stress the novelty.
Some examples are provided below (just to name a few):
https://www.comau.com/it/mate
http://www.skel-ex.com/
https://www.levitatetech.com/

Section 1.1, line 23: referring to the passive systems, the authors state that “commonly the mechanical design solutions are more complex”. Please extend this idea explaining its meaning in the text. Add references if necessary.

Section 1.2: the given title is “Industrial Exoskeletons: Control Solutions”, but the provided references referrer also to the rehabilitation field (e.g., [11]). I suggest removing the word “industrial” from the title.
In addition, some more refences can be added to enlarge the overview of the already studied solutions:
Secciani, N. et al. (2019). A novel application of a surface ElectroMyoGraphy-based control strategy for a hand exoskeleton system: A single-case study. International Journal of Advanced Robotic Systems.
Ryser, F. et al. (2017). Fully embedded myoelectric control for a wearable robotic hand orthosis. In 2017 IEEE International Conference on Rehabilitation Robotics (ICORR) (pp. 615-621).

In the same section, the SEA-based control should be at least mentioned (mainly because the proposed solution is based no the same control strategy). I suggest to add the following references:
M. Bianchi et al. Design of a Series Elastic Transmission for hand exoskeletons, Mechatronics, Volume 51, 2018, Pages 8-18.
N. Vitiello et al., "Functional Design of a Powered Elbow Orthosis Toward its Clinical Employment," in IEEE/ASME Transactions on Mechatronics, vol. 21, no. 4, pp. 1880-1891, Aug. 2016.

Section 2.2, line 96: “transparency requirements” are mentioned, but neither transparency issue nor the auto-alignment of the system are discussed in the paper. Please comment this point and, in case, remove the transparency requirement from the list.

Section 2.2.1, line 102: “D-H paramters2 should be D-H parameters (the same in other parts of the paper).

Figure 3: why did you put a 1-DOF constraints on the left of the first link of the shoulder? Could the R_s force be considered as a reaction of a rotational joint?

Section 2.2.2, lines 129-130: in the regulation systems (which are mentioned at the end of the paper), have some sizes been considered? Then, which are the maximum and minimum heights of the person wearing the exoskeleton?
Finally, how does the user’s weight influence the design of the system?

Equation 1: why the capstan equation has not been considered in the model? How does this neglection influence the torque calculation? Please, be specific.

Equation 2: have the friction forces been considered in the model? If yes, how could the authors calculate them?

Section 3.3, line 204: “To select the elastic belt, a shoulder equivalent stiffness of 200 Nm/rad”. Why the specific value of 200 Nm/rad has been chosen?

Figure 6: why, in the first graph, the slope changes at a certain point (and, consequently, the K_eq drops under 100 Nm/rad)?

Section 6: this section appears quite “disjointed” with respect to the rest of the paper, and it does not give additional importance to the contribution. In addition, it is not clear how the proposed SEA is mounted on the 3-DoF shoulder joint. I therefore suggest the authors to remove this section.

I really hope these suggestions could be useful to the authors to improve the quality and the effectiveness of their paper.

Yours faithfully,

The Reviewer

Author Response

Dear reviewer,

authors would like to thank you for your expertise and careful revision.

Please find in attached the response to your questions/comments.

Authors hope to satisfy your requests with the provided replies and revision.

Author Response File: Author Response.pdf

Round  2

Reviewer 3 Report

The revised paper clearly shows the detailed description of the state-of-the-art, simulative experiments and an improved presentation of associated conclusions. In terms of practical suitability, a closer look at the physical prototype and related experiments would be desirable but not necessary if practical results do not yet exist. Furthermore, the English language has extensively been revised in most points, but could be improved in terms of style and syntax.

Author Response

Considering the reviewer concern, the reply is as follows: the presented paper proposes the methodology for both mechanical and control design, describing in details the proposed approach, together with the details for the realization of the device. Authors agree that the experimental evaluation of the proposed exoskeleton + controller improves the results. However, since the main scope of the paper is the methodology description and the prototype is under realization, authors would highlight that the paper in the current form can be representative of the complete process of design and control of an industrial exoskeleton, also including a deep analysis of the state of the art solutions, making the contribution valuable.

Reviewer 4 Report

Authors have improved the presentation of the paper and have resolved concerns on the literature reviews. However, the experimental evaluation is required to demonstrate the validity of the proposed methods as well as the novelty of the work.

Reviewer 5 Report

The Authors have addressed all the points arisen by the reviewer.

In my opinion, the paper can be published in this new form.