Development and Implementation of an Anthropomorphic Underactuated Prosthesis with Adaptive Grip

Round 1

Reviewer 1 Report

The paper introduces a compliant prosthesis manufactured using 3D printing and tested.

The introduction lack on methodologies used for testing such thing. Papers should be mentioned and commented regarding this procedures that are important in medical field. These papers can be used and commented:

Chaparro-Rico, B.D.M., Martinello, K., Fucile, S., Cafolla, D. User-tailored orthosis design for 3d printing with plactive: A quick methodology (2021) Crystals, 11 (5), art. no. 561, . DOI: 10.3390/cryst11050561

Ceccarelli, M., Carbone, G., Cafolla, D., Wang, M. How to use 3D printing for feasibility check of mechanism design (2016) Advances in Intelligent Systems and Computing, 371, pp. 307-315. DOI: 10.1007/978-3-319-21290-6_31

-The mechanism detail must be improved, there are no equations kinematics or dynamics

-the control part must be improved with a detailed description of functioning related to mentioned equations
- Compliant test with quantitive results must be done using sensors such has force sensors or similar to test the compliance and the gripping force. Only photos are not enough. Plots and discussion are needed.

Figure 16 is very small and difficult to read.

All in all, the paper present potential but it must go deeper in to formal analysis with details on mechanism control and validation.

Author Response

Revisions Letter

Dear Reviewer,

We are writing to thank you for your response and your effort in reading and revising in detail our paper. We believe that all your suggestions have greatly improved the manuscript. You may find below your comments to the revision of the paper machines-1339980: “Development and implementation of an anthropomorphic underactuated”, and our answers.

We are also sending, separately, a new version of the paper with all changes highlighted in blue. If you have further questions we will be very glad to answer them. 

Comments about Technical remarks

Comment N°1

The paper introduces a compliant prosthesis manufactured using 3D printing and tested.

The introduction lack on methodologies used for testing such thing. Papers should be mentioned and commented regarding these procedures that are important in medical field. These papers can be used and commented:

[1] Chaparro-Rico, B.D.M., Martinello, K., Fucile, S., Cafolla, D. User-tailored orthosis design for 3d printing with plactive: A quick methodology (2021) Crystals, 11 (5), art. no. 561, . DOI: 10.3390/cryst11050561

[2] Ceccarelli, M., Carbone, G., Cafolla, D., Wang, M. How to use 3D printing for feasibility check of mechanism design (2016) Advances in Intelligent Systems and Computing, 371, pp. 307-315. DOI: 10.1007/978-3-319-21290-6_31

Answer

The reviewer's suggestion is sincerely appreciated. We have added the references and described them in section 1 “Introduction”. We also agree with the reviewer’s comment regarding the need to test the device, however we are at an initial stage of the prosthetic hands development. As stated in the last paragraph of the introduction, “The objective of the design is focused on minimizing the main causes of user discomfort, i.e., reducing the weight of the prosthesis and minimizing the number of signals required to control it.”

From our point of view, before testing the device the following steps must be addressed:

• Design the adaptive grasping mechanism.
• Image recognition control system.
• Integration of mechanical, control system and actuators

Currently we are working on the image recognition system which enables to determine with no external signals generated by the user (i.e. pressing buttons, electromiographic signals, etc.) the intended grip desired by the user.

We encourage the reviewer to watch the following videos that show ongoing progress currently not integrated in the prosthesis.

• https://youtu.be/_LUEIDgRges Here it can be seen how the camera distinguishes between the two hands, and only gives the order to grip the sphere when it touches the gloved hand, which resembles the prosthesis.
• https://youtu.be/sU-eRY4moaw In this video it can be seen that the algorithm only gives the order to grip when the sphere touches the "prosthesis" represented by the glove, when the sphere is close, but without contact, the order to "grip" is not given. It should be noted that the algorithm operates in real time.

Comment N°2

The mechanism detail must be improved, there are no equations kinematics or dynamics

Answer

As suggested by the reviewer, we have improved the mechanisms explanation including a description of the extension movement, the connection of the tendons to the drive pulleys and added pictures to help the understating of the text.

Considering that the purpose of this article is to prove the working principle of the proposed mechanism, we consider that addressing the kinematic and dynamic analysis is a subsequent step, which we are currently working on.

It is complex to address this analysis from the kinematic and dynamic equation perspective. On the one hand, by only using the kinematic equations it is difficult to address the contact force of the fingers with the object, which directly interfere with the prosthesis movements. By simultaneously using kinematics and dynamics, we find two difficult issues to consider:

• the friction between the tendon and the 3D printed parts.
• the elasticity of the cable.

Comment N°3

The control part must be improved with a detailed description of functioning related to mentioned equations

Answer

We truly appreciate the reviewer’s inquiry, control is a relevant issue in upper prosthetic limbs; however our work focuses on the conceptual design and the implementation/construction of the device. The control of the actuators was done manually as there are no sensors on the prosthetic hand.

Large deformations on the tendons occurs when a force is applied, thus it is not possible to apply an open loop position control by relating the drive pulleys rotation and the fingers position. Even without this problem, the differential mechanism is used to control several links with one motor, this allows to wrap the object making it impossible to correlate the pulleys rotation with the fingers position.

We are currently working on a closed loop control strategy for the prosthesis positioning system, the mechanism will be aided by cameras and an image recognition.

Comment N°4

Compliant test with quantitative results must be done using sensors such has force sensors or similar to test the compliance and the gripping force. Only photos are not enough. Plots and discussion are needed.

Answer

We understand the reviewers concern and we agree with his comment, nevertheless the goal of this work is to present a conceptual design and the implementation/construction of a prosthetic device with minimum weight and number of control signals required by the user. From this point of view, we analyze the number of actuators and presented from our perspective a mechanic solution (the differential mechanism) for the implementation of the adaptive grasping system.

As previously stated, current work is focused on the image recognition system for the implementation of a closed loop control. Testing the device is also an important issue, and the next step to continue the present research, nevertheless the global pandemic circumstances have made it difficult to carry out experimental activities generating delays in our investigation. Despite the above, I would like to assure the reviewer that we will address this issue *as soon as possible* in future publications.

Comment N°5

Figure 16 is very small and difficult to read.

Answer

We thank to reviewer for pointing out this issue. We have resized and rotated the image to improve readability.

Comment N°6

All in all, the paper present potential but it must go deeper into formal analysis with details on mechanism control and validation.

Answer

We are deeply grateful with your comments, we have improved the text according to your comments increasing the analysis level and we expect that the revised version meets your expectations.

We invite the reviewer to watch two short videos that show how the prosthesis works.:

• https://youtu.be/F8F93_5Q_so This video shows a 3kg weight lifting capability
• https://youtu.be/yb9zd77D-s8 This video shows a pincer grip, highlighting that the thumb executes the opposing movement first

Author Response File: Author Response.docx

Reviewer 2 Report

This paper describes the design of a prosthetic hand for wrist amputations. The designed hand uses three actuators, which are used to drive the movements of the little finger, annular finger and middle finger, of the index finger, and of the thumb, respectively. And, using the prototype hand, various grasping tasks were performed.

The authors carefully described the design method. Despite the authors’ efforts, there still remain many things to be improved, as follows:

1. Regarding the finger design, the author said that a finger can realize extension/flexion movements. In Figure 7, a single tendon for flexion movement is shown, and Figure 4 only describes the flexion movement. However, in this manuscript, the extension movement was realized using other tendon than the one used for the flexion movement, and it is explained that “the extension movement is controlled by a cable that passes through the channel located at the top.”. Here, it is difficult to understand because three motors were used in the system. I would like to add specific explanation in Figure 4.

- Describe Flexion/Extension movements, separately, and represent the movement of each tendon in detail.

- I could not find the explanation related to the top tendon anywhere in the manuscript.

2. In the thumb design, both movements, Extension/Flexion and Opposition/Reposition, (shown in Figure 7), are implemented using a single tendon and the distance difference between the path of tendon and the rotation axes. I have questions here.

- Are Extension/Flexion and Opposition/Reposition movements decoupled? Or, do they happen simultaneously?

- If two movements are decoupled, is the order of movements flexion -> opposition -> reposition -> extension?

- It's difficult to understand, so I hope you add a video about the movement.

3. To realize the opposition/reposition of the thumb, I think the tendon installed with an angle difference between the axis of rotation and the tendon path. However, the tendon path was located in parallel with the rotation axis, as shown in Figure 7. Is this a possible design to implement opposition/reposition movement?
4. In Figure 7, the tendon for index finger movement is made to pass through the thumb frame. And, I think the movement of the thumb affected the index finger motion. Is there no effect when controlling the index finger?

5. In this manuscript, three motors were used. Although I could understand the process of flexion movements in all fingers, please explain in detail how to implement extension motion.

6. Is position control possible?

7. In design of the prosthesis, it is difficult to fix the tendon with appropriate tension because of the limited design space. Please explain how you solved this problem.

8. There is no information about the used tendon. Please explain the material, diameter, and allowable stress of the used tendon.

9. There is no specification about the used actuator. Please explain the torque, rpm, and rated current.

10. How much is the payload of hook motion in Figure 14(b)?

11. I'm curious about the overall picture of the aluminum profile in Figure 11. What is the weight of the aluminum profile and how much is the gripping force?

12. Please explain how the movement of Figure 14(a) comes out using movements in Figure 5.

14. It would be nice to have a full picture including the actuators. Please explain the connection between the actuators and the tendon.
15. Table2 title figura -> figure
    
    
16. In Table2, the description of the mechanical differential is hard to see.

Overall, the description of the design is too short, and the description of the figure is too cursory. In the revised one, the description should be much more specific and in detail.  

I hope my comments are helpful to improve your paper.

Author Response

Revisions Letter

Dear Reviewer,

We are writing to thank you for your response and your effort in reading and revising in detail our paper. We believe that all your suggestions have greatly improved the manuscript. You may find below your comments to the revision of the paper machines-1339980: “Development and implementation of an anthropomorphic underactuated”, and our answers.

We are also sending, separately, a new version of the paper with all changes highlighted in blue. If you have further questions, we will be glad to answer them.

Comments about Technical remarks

Comment N°1

1. Regarding the finger design, the author said that a finger can realize extension/flexion movements. In Figure 7, a single tendon for flexion movement is shown, and Figure 4 only describes the flexion movement. However, in this manuscript, the extension movement was realized using other tendon than the one used for the flexion movement, and it is explained that “the extension movement is controlled by a cable that passes through the channel located at the top.”. Here, it is difficult to understand because three motors were used in the system. I would like to add specific explanation in Figure 4.

- Describe Flexion/Extension movements, separately, and represent the movement of each tendon in detail.

- I could not find the explanation related to the top tendon anywhere in the manuscript.

Answer

The reviewer's suggestion is sincerely appreciated, there was no description for the extension movement. This was done intentionally because this part of the design has no additional complexity, same as in the flexion movement. Additionally, it is not used for any of the hand grips described in figure 3, where the emphasis is done on the closing motion. However, we agree with the reviewer’s comment and the text has been modified in section 4.3 to include a description of the extension movement. Additionally, as the reviewers stated the difficulty to understand how it is possible to use only three motors in the system, figure 4 was modified to include the motor drive pulley setup, along with a description of the extension and flexion movements. 

Comment N°2

In the thumb design, both movements, Extension/Flexion and Opposition/Reposition, (shown in Figure 7), are implemented using a single tendon and the distance difference between the path of tendon and the rotation axes. I have questions here.

- Are Extension/Flexion and Opposition/Reposition movements decoupled? Or, do they happen simultaneously?

- If two movements are decoupled, is the order of movements flexion -> opposition -> reposition -> extension?

- It's difficult to understand, so I hope you add a video about the movement.

Answer

According to the reviewer suggestion, the movements are decoupled and performed in sequence to improve the gripping capabilities of the prosthetic hand. The order of movement is opposition → flexion for grasping, and extension → reposition for opening, this sequence is explained in the text as follows:

“The execution of these movements must be performed sequentially as shown in Figure 6. The movement starts with the opposition (Figure 6a), performing a 90°rotation relative to the red axis, followed by the flexion movement represented in Figure 6b, reaching the final position in Figure 6c. This sequence of movements is achieved in the same manner as in triphalangeal fingers, i.e. by adjusting the distance between the axis of rotation and the tendon.”

The opening movement is not explained, mainly because it is not used for any of the hand grips described in figure 3, and thus it is not necessary to perform a specific sequence. However, due to the reviewer’s comment, we modified the text in section 4.4 accordingly.

We share with the reviewer two short videos in which the prosthesis can be seen in action.

• https://youtu.be/F8F93_5Q_so This video shows a 3kg weight lifting capability.
• https://youtu.be/yb9zd77D-s8 This video shows a pincer grip, highlighting that the thumb executes the opposing movement first.

Comment N°3

To realize the opposition/reposition of the thumb, I think the tendon installed with an angle difference between the axis of rotation and the tendon path. However, the tendon path was located in parallel with the rotation axis, as shown in Figure 7. Is this a possible design to implement opposition/reposition movement?

Answer

We are sincerely grateful to the reviewer for pointing out this problem generated by switching from a two dimensional to a three dimensional image. We have added the figure 7 to describe in detail the path of the tendons. Additionally, an explanatory text is incorporated in section 4.4 to describe the figures and tendon paths.

Comment N°4

In Figure 7, the tendon for index finger movement is made to pass through the thumb frame. And, I think the movement of the thumb affected the index finger motion. Is there no effect when controlling the index finger?

Answer

As mentioned in comment N°3 we made a mistake in figure 7 and we thank the reviewer for detecting the problem. The tendon for the index finger passes through a hollow shaft located behind the thumb. We replaced figure 4 to show this detail and added figure 8

Comment N°5

In this manuscript, three motors were used. Although I could understand the process of flexion movements in all fingers, please explain in detail how to implement extension motion.

Answer

We acknowledge the reviewer’s concern. As mentioned in a previous answer (comment N°1), the flexion/extension movement and its connection to the driving pulley is now described in a new paragraph of section 4.3. Figure 4 has been modified to include the drive pulley and its connection to the tendons, we sincerely hope that explanation given is clear.

Comment N°6

Is position control possible?

Answer

Although we have no information on the cable material, the trials show large deformations on the line when a force is applied, thus it is not possible to apply an open loop position control between finger and the pulley. Even without this problem, the mechanism is used to control several links with one motor and allowing to wrap the object, this makes it impossible to correlate the pulleys rotation with the fingers position.

We are currently working on a closed loop control strategy for the prosthesis positioning system, the mechanism will be aided by cameras and an image recognition.

We encourage the reviewer to watch the following videos that show ongoing progress currently not integrated in the prosthesis

• https://youtu.be/_LUEIDgRges Here it can be seen how the camera distinguishes between the two hands, and only gives the order to grip the sphere when it touches the gloved hand, which resembles the prosthesis.
• https://youtu.be/sU-eRY4moaw In this video the algorithm only gives the order to grip when the sphere touches the "prosthesis" represented by the glove, when the sphere is close, but without contact, the order to "grip" is not given. It should be noted that the algorithm operates in real time.

Comment N°7

In design of the prosthesis, it is difficult to fix the tendon with appropriate tension because of the limited design space. Please explain how you solved this problem.

Answer

The reviewer is right, the assemble process was difficult and required much patience to put all strings through the channels with very limited space. The complete sequence of assemble is difficult to describe, sub-assemblies are made and later integrate to the prosthesis. Subsequently the two ends of each cable bust be secured. We begin with tying knots in the distal section of the hand, at this stage we can adjust the correct length of all the tendons enabling the desired movement response.

When this is complete, the tendons must be connected to the drive pulleys, the first step is to secure the flexion tendons to the pulley, then we energize the actuators and apply tension to the flexion tendons allowing to tie knots in the extension tendons more easily.

We hope this information is useful for the reviewer, this section it is difficult to explain in depth, we hope that the mentioned sequence might give you an insight of the assembly procedure

Comment N°8

There is no information about the used tendon. Please explain the material, diameter, and allowable stress of the used tendon.

Answer

We appreciate the observation made by the reviewer, the tendon is made of a monofilament strand with a diameter of 0.5 mm, the manufacturer does not mention the material or the allowable stress. The maximum force supported is rated at 14,6 kg. When considering the section area, the calculated tensile strength is 730 Mpa. This information is now incorporated in the text, excluding the calculated tensile strength.

Comment N°9

There is no specification about the used actuator. Please explain the torque, rpm, and rated current.

Answer

The actuators are servomotors, the following table is now included in the text, including the requested specifications:

Servomotors were driven at 5V.

According with the reviewer’s suggestion we have included this information in table 6 located in section 4.2.

Comment N°10

How much is the payload of hook motion in Figure 14(b)?

Answer

We appreciate that the reviewer has taken the time to analyze the details of the presented device. We tested the capacity to lift weight with the hand, it was possible to hold 3 kg, with more weight added the tendons broke. This information is not presented in the text as we do not account for specialized instrumentation to perform this kind of measurements. The paper focuses on the design and implementation/construction of the proposed device.

https://youtu.be/F8F93_5Q_so  This video shows a 3kg weight lifting capability

Comment N°11

I'm curious about the overall picture of the aluminum profile in Figure 11. What is the weight of the aluminum profile and how much is the gripping force?

Answer

The aluminum profile is a  V-Slot linear rail 40x40 mm and 250 mm long, its weight is approximately 350 grams. Regarding the gripping force, we do not account with adequate instruments to perform these measurements, thus we did not include this information and focused on the dexterity of the design. Further tests will be performed in future work

Comment N°12

Please explain how the movement of Figure 14(a) comes out using movements in Figure 5.

Answer

Figure 14a) corresponds to a gravity grip achieved by the extension of the five fingers, as the hand does not need to wrap an object, the opening sequence is not relevant only the final position. Figure 5 shows the closure sequence in the flexion process, as the extension movement was not explained in the text, we are assuming it could have generated confusion.

Comment N°13

It would be nice to have a full picture including the actuators. Please explain the connection between the actuators and the tendon.

Answer

We thank to the reviewer for the interest and dedication required to review in depth our work. As requested, we have included some pictures with the full device and some pictures with details of the pulleys. If the editor and the reviewer consider it appropriate, we can include the pictures in the manuscript.

Comment N°14

Table2 title figura -> figure

Answer

We thank the reviewer for pointing out this issue. We have revised the document and fixed the problem.

Comment N°15

In Table2, the description of the mechanical differential is hard to see.

Answer

We thank to reviewer for pointing out this issue. We have revised the document and fixed this issue adding colors to the image.

Comment N°16

Overall, the description of the design is too short, and the description of the figure is too cursory. In the revised one, the description should be much more specific and in detail. I hope my comments are helpful to improve your paper

Answer

We are grateful for your comments, we believe the text has been greatly improved, we hope that the revised version meets your expectations.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Authors addressed my points explaining why some points cannot be addressed completely.. I suggest to go deeper in the points I suggested in future works to strenghten your proposal.

Author Response

Revision Letter

Dear Reviewer,

The Authors would like to thank the Reviewer for taking the time to go over our manuscript one more time, we appreciate the comments submitted. As expected, all comments were properly considered by the Authors, which led to minor manuscript adjustments, enhancing the quality and clarity of the text. Furthermore, all the questions submitted by the Reviewer are answered and displayed in blue.

We are also sending, separately, a new version of the paper with all changes highlighted in blue. If you have further questions we will be glad to answer them.

Comments about Technical remarks

Authors addressed my points explaining why some points cannot be addressed completely. I suggest to go deeper in the points I suggested in future works to strengthen your proposal.

We are sincerely grateful with the appreciations indicated by the reviewer. As it was commented in round N°1, we deeply agree with the comments received. The developments currently in progress cover these points because we envision that they will strengthen the progress of the research. Evidence of this preliminary development was shown in the videos submitted during round 1.

We have included in the conclusions a section with comments regarding the direction of this research and future challenges this work will cover. These guidelines are in agreement with the reviewer's comments.

We really appreciate all the time you have dedicated to this paper, both your comments on the first and second revisions have helped to substantially improve the document and regardless of the outcome we are grateful for that.

Author Response File: Author Response.docx

Reviewer 2 Report

Although authors have tried to address issues raised by reviewers, it still has several unclear points on its main contribution.

The most critical one is its structural rigidity. From the figure 4, it seems that it keeps its current pose only by tension of the tendon. Another point in this figure is that there is no precise mechanical element on the joints. Are there any bearings or all the joints are just friction hinge. If it is just a friction hinge, isn't any problem of rigidity for manipulation?

Of course, the rigidity may not be an issue depending on its purpose. If so, please clarify that point in the introduction.

Overall, first clarify its purpose, and add more rigorous reviews on the related prosthetic hand.

Author Response

Revision Letter

Dear Reviewer,

The Authors would like to thank the Reviewer for taking the time to go over our manuscript one more time, we appreciate the comments submitted. As expected, all comments were properly considered by the Authors, leading to minor manuscript adjustments, enhancing the quality and clarity of the text. Furthermore, all the questions submitted by the Reviewer are answered and displayed in blue.

We are also sending, separately, a new version of the paper with all changes highlighted in blue. If you have further questions, we will be glad to answer them.

Comments about Technical remarks

Although authors have tried to address issues raised by reviewers, it still has several unclear points on its main contribution.

The most critical one is its structural rigidity. From the figure 4, it seems that it keeps its current pose only by tension of the tendon. Another point in this figure is that there is no precise mechanical element on the joints. Are there any bearings or all the joints are just friction hinge. If it is just a friction hinge, isn't any problem of rigidity for manipulation?

As shown in the video previously shared (https://youtu.be/F8F93_5Q_so), the device can lift 3 kg, after the test no damage was observed in the tendons, and it was possible to keep testing the device, from our perspective there will be no rigidity problems when manipulating common objects, of course for heavier objects the reviewers concern can be a problem. We have recently acquired a strand of kevlar with 2760 Mpa reported tensile strength, almost 4 times the strength of the fishing wire previously used.  This will allow the prosthesis to lift heavier objects. It is also possible to use a thicker fishing line (bigger diameter), commercially it is possible to find it up to 2mm wire capable of resisting 83 kg of tension 5,7 times more than what is being used. If we use stronger tendons we must verify that all 3d printed parts are able to resist the added stress, it is preferred that the tendon is the weakest link in the chain. Considering that a spare tendon will be cheaper and easier to obtain..

The reviewer is right, there are no special elements (bearings) in the joints, the proposed design did not generate any problems during the tests executed, it allowed to maintain a simpler mechanism. On the other hand, the device did not show wear when performing repetitive movements.

Of course, the rigidity may not be an issue depending on its purpose. If so, please clarify that point in the introduction.

The purpose of our research is to develop a low cost prosthetic hand that can be easily manufactured and allows 

for the execution of common tasks with ease of control. We have described our vision and there are several previous steps to accomplish before succeeding. For the present work our goal was to develop, manufacture and test the adaptive grip mechanism achieving several grips, as depicted in figures 9 to 16.

Originally this was written in the text as “ The objective of the design is focused on minimizing the main causes of user discomfort, i.e., reducing the weight of the prosthesis and minimizing the number of signals required to control it. The problem is addressed by the design, and manufacture of an adaptive grip mechanism.”

Overall, first clarify its purpose, and add more rigorous reviews on the related prosthetic hand.

We thank the reviewer for pointing out this issue, we have updated the search for information and included 7 new publications to improve the review of the latest advances in the development of hand prostheses. We have also included a text in the introduction to clarify the purpose of the study.

We really appreciate all the time you have dedicated to this paper, both your comments on the first and second revisions have helped to substantially improve the document and regardless of the outcome we are grateful for that.