Design of Tendon-Driven Mechanism Using Geometrical Condition

Round 1

Reviewer 1 Report

The review report is attached.

Comments for author File: Comments.pdf

Author Response

Thank you for suggestive review. The comments were very important and I believe the paper is revised to be more attractive to the reader. Here is my answer to your comments and question. Thank you again.

The paper is written well overall.

--> I’m very glad to seeing your comments. I believe the paper contributes the design of the tendon-driven robot.

1. Figure 1 is not explicit or clear in representing the positions of the pulley. The reviewer suggest to explicitly draw the pulleys and tendons so that it is easier to understand the driving mechanism. For instance, draw the pulley as a circle that is rigidly attached to the links and then draw the tendon that loop around the contour of the pulley. In addition, it is difficult to understand the following description: “In this paper, the position of the attaching point P0 did not move when the actuator pulled the wire, although the actual linear actuator does move while pulling the wire.”

--> Thank you for good suggestion. I modified the picture of Figure 1(a) to show that the wire turns around the pulley. Additionally, the sentence that the point P0 attaching on the rod of the linear actuator did not move is revised. The paper discusses just a moment when the actuator drove and the moving distance of the point is very short. Therefore, it is assumed that the position of the P0 did not move on the reference flame

2. Are there any constraints on the positions of the pulleys? For instance, to limit the size of the physical construction of each link, the pulley shall not be too far away from the link. In addition, how are the constraints incorporated into the proposed geometrical approach?

--> Because the size of the link is limited according to the size of the robot, the pulley should be set within the size of the link. Another physical constraint is that some sets of the pulley position leads contacting on the link. These are explained in the conclusion.

3. In Figure 4, please denote the path of the tendon and the locations of the pulleys. In addition, detailed descriptions may be provided in the caption of Figure 4 to make the figure self-explanatory.

--> I attach the picture that shows the tendon at loosing and position of the pulley. It will be comprehensible to the reader.

4. The purpose of the leg design is to achieve vertical force at the toe for vertical jumping. Has the gravitational force of the leg been considered in the proposed approach? For vertical jumping, the vertical force at the toe may not be colinear with the gravitational force of the leg, which may generate a moment to the whole leg and eventually a spin motion to the leg.

--> The main aim of the paper is to generate the desired force on the end effector. The experiment of the vertical jumping is one of the applications. Because the jumping motion is affected by inertia of the link and mass as well as the force, we had an assumption that the link is very light and the inertia does not affect the jumping. Because the physical link is constructed by light material (PLA and ABS), it is assumed that the inertia is negligible. In the jumping experiment, we confirmed that the robot jumped vertically by the appropriate position of the pulley. For the general robot model, the inertia should be considered. It is explained in the conclusion. Thank you for pointing out the important issue.

5. The pulleys will result in friction to the tendon and thus the proximal force and distal force of the tendon will be different. How significant is the friction? A reference with friction formulation of tendon-driven mechanism is suggested: Lai et al., 2019, An Integrated Sensor-Model Approach for Haptic Feedback of Flexible Endoscopic Robots; Li et al., 2019, Distal-end force prediction of tendon-sheath mechanisms for flexible endoscopic surgical robots using deep learning.

--> Thank you for introducing very important papers. I checked and refer the paper. The friction should be very important factor for the tension transmission. Based on the proposed procedure, more general model including the friction will be discussed in future. The sentence is added in the conclusion.

6. The authors may also want to add a video clip with the jumping motion of the robotic leg to demonstrate the result.

--> I made video clip in which the robot succeeded and failed in jumping. Thank you for good suggestion.

7. Are there any limitations of the approach? If so, it will be good to add in the conclusion section.

--> The items 2 and 5 you pointed out are the limitation of the proposed method. Thank you for providing the opportunity to think about our research. I wrote the sentence in the conclusion.

Some grammar errors exist in the some sentences, for instance,

“A tendon-driven robot has an advantage of setting mass distribution and facilitation of

the motion and so on.”

“In order to setting pulley position for the desired direction and magnitude of

force, a geometrical condition to realize the desired motion is adopted.”

--> Thank you for pointing out the grammar errors. I revised the expression.

Reviewer 2 Report

The paper developed a model for designing tendon-driven robot with one actuator. In particular, the work was concerned with driving a condition through which the allocation of pulley can allow a desired force magnitude and direction. A preliminary physical robot was built and experimented and results were compared with the model predictions. Generally, the paper is well written and organised. It should prove useful for researchers interested in developing such type of robots. My comments aim to improve the clarity of the work at a number of places:

1. Line 8> Please avoid words such as “so on” as they are not typically used within scientific documents
2. There is excessive use for the word “we” throughout the paper. Whilst not unacceptable, I think using “I” or trying to avoid “we” would make better sense
3. I believe that the introduction can benefit significantly from a discussion of jumping performance in real jumpers (animals and insects). At the moment the introduction is very limited to robotic platforms and I think adding a paragraph or two on the jumping characteristics of natural jumpers will allow a better global picture of the jumping problem. For example, the study “Energy and time optimal trajectories in exploratory jumps of the spider Phidippus regius. Scientific reports, 8(1), 7142.” compares jumping performance of almost all reported jumping insects and I suggest highlighting the main outcomes from this here.
4. Figure 1a: I am not sure why the wire between P0 and P1 passes by J1. It can be confusing and hence maybe improve the clarity of the figure in this respect
5. Equation 1: check the force components are correctly written, i.e. F_y
6. Lines 104-105> please discuss the implications/limitations from the adopted assumption
7. Figure 2: Improve the size of the fonts used, particularly in fig 2b – at the moment the look so big and inconsistent
8. Line 158> correct l4 value
9. Figure 4> Is it possible to add an exploded view of the robot to see the different parts involved?
10. Line 168-169> please include all information about strain gauge and force scale including model number, manufacturer, resolution, etc.
11. Line 181> “is occurred” - correct grammar
12. Line 216> I think you mean Table 2
13. Table 2> It would be useful to add the percentage error/difference between measured and desired values and comment on the significance and reason for this discrepancy
14. Line 229 and 231> “was occurred” check grammar
15. Figure 7> please include a length scale
16. Lines 263-265> not clear to me how this was achieved, please elaborate on the method and why this measurement procedure?
17. Table 3> please add a column to evaluate percentage error/difference between measured and desired values and comment on the significance and reason for this discrepancy
18. It would be useful to comment on any scalability issues of the proposed method, i.e. are there scales/sizes of robots that this method would be better or worse suited for?

Author Response

Thank you for fruitful comments and questions. I revised the paper. I believe the paper gets attractive for the reader. Here is my answer to your comments and question. Thank you again for reviewing and pointing out very important issues.

Line 8> Please avoid words such as “so on” as they are not typically used within scientific documents

There is excessive use for the word “we” throughout the paper. Whilst not unacceptable, I think using “I” or trying to avoid “we” would make better sense

--> Thank you for pointing out the issues. I change the expression without using “so on”. Also, I modified some sentences to avoid using “we”.

I believe that the introduction can benefit significantly from a discussion of jumping performance in real jumpers (animals and insects). At the moment the introduction is very limited to robotic platforms and I think adding a paragraph or two on the jumping characteristics of natural jumpers will allow a better global picture of the jumping problem. For example, the study “Energy and time optimal trajectories in exploratory jumps of the spider Phidippus regius. Scientific reports, 8(1), 7142.” compares jumping performance of almost all reported jumping insects and I suggest highlighting the main outcomes from this here.

--> Thank you for providing important information. The jumping motion is one of the powerful examples, and it is important to introduce studies about animal and insect jumping. I add a paragraph in Section 3.1 referring the paper you mentioned.

Figure 1a: I am not sure why the wire between P0 and P1 passes by J1. It can be confusing and hence maybe improve the clarity of the figure in this respect

--> As you pointed out, Figure 1a will confuse the reader. I revised the figure in which the wire does not passes by J1. Thank you.

Equation 1: check the force components are correctly written, i.e. F_y

--> Thank you for informing the typo. I revised.

Lines 104-105> please discuss the implications/limitations from the adopted assumption

--> The pulley radius described in Higashimori’s paper was arbitral. On the other hand, we adopted the pulley whose radius is negligibly small. This is one of the limitation of our proposed method. A wrote the sentence about it in the conclusion.

Figure 2: Improve the size of the fonts used, particularly in fig 2b ? at the moment the look so big and inconsistent

--> Because the picture is enlarged, the font is also enlarged. I adjust the picture. Thank you.

Line 158> correct l4 value

--> I correct the value. Thank you.

Figure 4> Is it possible to add an exploded view of the robot to see the different parts involved?

--> I attach the picture from another angle. I think the shape of the parts is comprehensible.

Line 168-169> please include all information about strain gauge and force scale including model number, manufacturer, resolution, etc.

--> The expression of the the information about the strain gauge is added. Also, the resolution of the micro controller that measure the nonlinear resistance to the load is explained. Thank you for asking about it.

Line 181> “is occurred” - correct grammar

--> Thank you for pointing out the grammar error. I revised.

Line 216> I think you mean Table 2

--> I correct the number. Thank you.

Table 2> It would be useful to add the percentage error/difference between measured and desired values and comment on the significance and reason for this discrepancy

--> It is very important suggestion. The relative error is added.

Line 229 and 231> “was occurred” check grammar

--> I corrected the grammar errors. Thank you.

Figure 7> please include a length scale

--> By using the information of the length of the link, it is possible to calculate the moving distance of the knee joint. The expression of the moving distance is explained in Section

Lines 263-265> not clear to me how this was achieved, please elaborate on the method and why this measurement procedure?

--> The procedure to determine the position of the pulley is added. The desired forces Fx and Fy are set, and then the pulley position is calculated. The tangential angle xi is not set but calculated by xi = tan^{-1}(Fx/Fy).

Table 3> please add a column to evaluate percentage error/difference between measured and desired values and comment on the significance and reason for this discrepancy

--> Thank you for good suggestion. Similar with Table 2, the relative error is added.

It would be useful to comment on any scalability issues of the proposed method, i.e. are there scales/sizes of robots that this method would be better or worse suited for?

--> I expect this procedure will be adopted for arbitral size. However, considering that the radius Ri from the joint depends on the magnitude and direction of the force as well as robot posture, some combination of the magnitude and direction will provide moderate radius by which pulley can be placed within the link. On the other hand, other combination will provide larger value by which the pulley is placed far from the link. It will be discussed in future.

Round 2

Reviewer 1 Report

The author has done a good job revising the manuscript. I have no more comments on the manuscript, but the author may want to add some texts in the video clip to explain why one jumping case is successful and the other failed.

Author Response

Thank you for good suggestion. I add some descriptions in the video. Thank you again.