Structural Optimization Method of a FinRay Finger for the Best Wrapping of Object

Round 1

Reviewer 1 Report

The authors propose a new and simple way to mathematically evaluate the wrapping of the finger around the object.  The reviewer agrees to the need for a new evaluation method for the Fin Ray gripper. 

The key feature of the Fin Ray gripper is the gripping and manipulation of objects of various shapes. However, the proposed method has not been shown the other cases about objects of various shapes. Furthermore, the reviewer thinks that the hysteresis characteristics of the finger are very important for determining its structure, but the reviewer could not find this experiment result.

Therefore, the reviewer thinks that this manuscript should be improved to publish.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

In this paper Suder et al described optimization of FinRay finger structure for wrapping of objects. Such fingers are important for grippers to grab irregular objects optimally, while benefits of them is simple passive design, which could be combined with both hard and soft actuators.

Strength: Methodology is systematic, carefully describing the work and different characterizations, complemented with mechanical modelling. Obtained data is well reported.

Weakness: Authors do not show any demonstration how this optimization would relate to any real application, for example simple grabbing/holding performance. Just (simple) tests/demonstrations of few fingers would greatly enhance the article.

Another important point missing is visual summary. It is easy to get lost in many different graphs and tables presented. Though they provide useful values when implementing particular finger, the “take home message” should be more clearly illustrated, to train the reader´s intuition about them. For example, deflection coefficient has been used throughout the paper as main characteristic of the structures. There should be chart (e.g. bar chart) showing all deflection coefficient value, these bars could be grouped and listing key parameters, such that structures could be compared. So all trends could be much easily seen!

How well prior arts have been cited: Numerous other interesting works of Soft FinRays have been shown for robotics, some of which also demonstrate mathematical modelling. I believe it would be important to cite more thoroughly these prior works. Some few such examples of references to these works are:

https://ieeexplore.ieee.org/abstract/document/9116052

https://www.frontiersin.org/articles/10.3389/frobt.2016.00070/full

https://www.daaam.info/Downloads/Pdfs/proceedings/proceedings_2011/1247_Pfaff.pdf

https://ieeexplore.ieee.org/document/8813388

Overall articles is well written. Language is simple, clear and easy to read.

Recommendation is to publish after revisions listed in section weaknesses ( 1)Some realistic test example and 2) summary chart)

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 3 Report

the paper is well written and provides some interesting investigations into the wrapping of fin ray fingers via simulations and some experimental validation. several aspects of the paper need more clarification on the reasoning behind the chosen approach and better supporting evidence to support claims made. more specific feedback below:

1. minor typo in introduction "press-ing"
2. quantifying the ability of a fin ray finger to wrap around an object using the Yend and Ymax is clear and has been utilised in a relatively similar way in previous work, yet the benefit of summing the ratio between those two over each mm of compression was not clear to me. I assume calculating this same ratio just at maximum compression should yield the same conclusions, as this is where the maximum adaptation of the fingers is most evident
3. a statement explaining why 8 mm specifically was chosen as the maximum compression is needed
4. the authors present several filling patters in figure 2, which includes some interesting alterations involving branching, yet no explanation was given as to how and why those designs parameters specifically were chosen to be studied? why not include parameters such as angle and spacing of internal ribs that were studied in previous work? this way a good benchmark could be established based on relevant previous design modifications in the literature
5. a reasonable variation in the values of the chosen design parameters were included in the studied in increments of the nozzle diameter, but it would have been useful to also consider changing size/shape of the test object to confirm if this is affecting the results. later in the paper the object approach is considered but overall the methodology of testing is not very clear
6. for the real testing, I assume the forces are measured using a force/torque sensor mounted at the end effector rather than force sensing at robot's joints. if so the name and specs of this sensor should be stated
7. the reasoning behind the real tests was not clear. why those set of tests in particular were chosen? an introduction explaining the objectives of the experiments and the methodology followed is needed
8. my understanding the plots in the results section are for the 5N force test rather than the 8 mm compression test. if that's the case then where are the results from the 8mm compression and how do those two different modes of testing compare? some explanation on this before presenting the results would help the reader
9. also, more information about the simulation setup before showing the results would be helpful.
10. one suggestion would be to use statistical methods such as ANOVA to identify the most significant design factors and focus on presenting the plots relating to those as currently there are too many figures in the paper
11. figure 14 is confusing to interpret. would be useful to only present the simulation results that were conducted in the real experiment so the match between simulation and real tests is clearer
12. in the final conclusion, the authors claim that fingers without any internal structure (arguably no longer a fin ray design) gives the greatest wrapping. However, this conclusion is not well supported as it is based on the tested design variations the authors chose to address in this paper, but might not be necessarily the case when including other design parameters (as explained in point 3). Additionally, a finger with no internal structure to propagate forces will probably struggle to find real applications without sufficient contact forces to maintain a stable grasp. Such limitations need to be mentioned and ideally tested to benchmark the force reduction compared to the original festo design.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

1. The other related works (such as electroadhesive method, magnetorheological elastomer)  should be included in Introduction.
2. It is better to insert cyclic hysteresis result.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 3

Reviewer 1 Report

Most of Reviewer's comments were reflected. Reviewer thinks that this manuscript is enough to be published in this journal.