Study and Design of a Machine Learning-Enabled Laser-Based Sensor for Pure and Sea Water Determination Using COMSOL Multiphysics

Round 1

Reviewer 1 Report

·        In this research author proposed sa novel method for real-time detection of label-free biochemical of salty water by combining various optics concepts with a machine learning system.

·        Abstract is strongly needed to improve. The last sentence in abstract part shows that contribution is very weak or experimental results are no complete. Please re-write and improve the abstract. Apart from this there are large number of spelling mistakes and grammatical issues. Please read it thoroughly. For example in the second line of the abstract “combinng” is a spelling mistake and similar for whole article.

·        Introduction section is very short. It is not covering the whole definition of an introduction. Please extend as per your work problem.

·        Contribution section is missing in the end of introduction section. Add a contribution paragraph as a second last paragraph in introduction section.

·        Paper organization section is missing in the end of introduction of section. Briefly describe the section and subsection of your whole menu script in one paragraph. Add this paragraph in the end of the introduction section.

·        There is no literature review section in this research. In order to prove that your research is beyond state-of-the-art and experiment results are more effective, you have needed to add a literature review section must. Author should present a table of comparison of previous published articles with this manuscript and show the importance of this menuscript why it should be consider for publication although there are already many good solutions have been developed by researchers.

·        What are the challenges and gaps of research? Author should explain challenges and gaps in this area of research in a separate section.

·        What are the limitations of the research? Have you compared your results with already published articles? If yes! Then present it in tabular form.

·        Try to avoid heading after heading such as:

§  Concept and methodology

§  2.1. Sensor concept and design

·        Add some text between all such headings and subheadings.

·        Author needs to provide the source code in order to prove that obtained results are correct and valid. As the whole research looks like an imaginary concept.

 

·        However, technically and practically paper is good. I would like to suggest this article for publication after major changes. The article is not acceptable in present form. 

Author Response

Response to Reviewer 1

 

The authors would like to take this opportunity to thank the reviewer for his time, consideration and insightful feedback. All the reviewer’s comments and suggestions have been taken into consideration and addressed in the revised version of the paper. All changes have been highlighted in yellow in the revised manuscript.

 

• Abstract is strongly needed to improve. The last sentence in abstract part shows that contribution is very weak or experimental results are no complete. Please re-write and improve the abstract. Apart from this there are large number of spelling mistakes and grammatical issues. Please read it thoroughly. For example, in the second line of the abstract “combinng” is a spelling mistake and similar for whole article

The authors would like to thank the reviewer for his valued comments. The abstract has been updated as per the reviewer’s suggestions.

In this work, the proposed sensor has been designed and validated using COMSOL Software. There is no fabrication involved at this stage.

All spelling and grammatical mistakes have been fixed.

• Introduction section is very short. It is not covering the whole definition of an introduction. Please extend as per your work problem.

The authors would like to thank the reviewer for their valued comment. The introduction section has been edited and different paragraphs have been added, including the contribution paragraph and the paper organization part.

• Contribution section is missing in the end of introduction section. Add a contribution paragraph as a second last paragraph in introduction section.

The authors would like to thank the reviewer for their valued comment. The contribution paragraph has been added to the introduction as recommended.

• Paper organization section is missing in the end of introduction of section. Briefly describe the section and subsection of your whole menu script in one paragraph. Add this paragraph in the end of the introduction section.

The authors would like to thank the reviewer for their valued comment. The paper organization part has been added to the introduction as suggested.

• There is no literature review section in this research. In order to prove that your research is beyond state-of-the-art and experiment results are more effective, you have needed to add a literature review section must. Author should present a table of comparison of previous published articles with this manuscript and show the importance of this manuscript why it should be considered for publication although there are already many good solutions have been developed by researchers.

The authors thank the reviewer for their valued comments. A section listing the different techniques employed for sensing pure and salty water has been added. The advantages and disadvantages have been presented as well. A comparison table has been added to highlight our research contribution and its advantage towards other techniques. 

• What are the challenges and gaps of research? Author should explain challenges and gaps in this area of research in a separate section.

The authors thank the reviewer for their feedback. The disadvantages and limitations of techniques used for sweater sensors’ development have been detailed in the paper. A common challenge facing researchers is the difficulty of collecting real data to tune the ML model.

• What are the limitations of the research? Have you compared your results with already published articles? If yes! Then present it in tabular form.

The authors thank the reviewer for their valued comments. As mentioned in a previous point, a section listing the different techniques employed for sensing pure and salty water, and their limitations, have been added.

 

• Try to avoid heading after heading such as:
• Concept and methodology
• 2.1. Sensor concept and design

Add some text between all such headings and subheadings.

The authors would like to thank the reviewer for their valued remark. Text has been added between headings and subheadings.

• Author needs to provide the source code in order to prove that obtained results are correct and valid. As the whole research looks like an imaginary concept.

The authors thank the reviewer for their valued comment. The source code is submitted with response to the reviewers’ comments.

• However, technically and practically paper is good. I would like to suggest this article for publication after major changes. The article is not acceptable in present form. 

The authors thank the reviewer for his valued feedback. All comments have been addressed as suggested.

 

 

 

 

Author Response File: Author Response.docx

Reviewer 2 Report

The title of the article does not correspond to its actual content.

The purpose of the study is not clear.

A very weak analysis of the results obtained.

The scientific nature of the article is doubtful. It deals with the concept of building a system for determining the salinity of water, which is closer to production than to science. There are no experiments confirming the calculations. There is no justification for the choice of sensor components. 

There is no line numbering in the text of the manuscript.

Questions and comments on the text of the manuscript:

1. There is no comparative analysis of water methods and sensors in the Introduction section. What is worse than other methods of water analysis? There is no discussion of other devices.

2. The explanation in two paragraphs at the bottom of page 2 is elementary physics, corresponding to the level of a school textbook, not a scientific journal.

3. Figure 2 needs correction: reduce the range on the ordinate axis, and on the abscissa axis you need to present a uniform scale, not a histogram.

4. The criteria for choosing the size of domain1 and domain 2 are not specified.

5. Figure 3 is also presented incorrectly along the abscissa axis.

6. Some paragraphs represent well-known information (for example, the last paragraph of page 6).

7. The criterion for determining whether the water is clean or salty is not specified if it was modeled only for S=0 and S=35.

Author Response

Response to Reviewer 2

 

The authors would like to take this opportunity to thank the reviewer for his time, consideration and insightful feedback. All the reviewer’s comments and suggestions have been taken into consideration and addressed in the revised version of the paper. All changes have been highlighted in yellow in the revised manuscript.

 

• The title of the article does not correspond to its actual content.

The authors would like to thank the reviewer for their valued comments. The title has been changed to better reflect its content.

• The purpose of the study is not clear.

The authors would like to thank the reviewer for their valued feedback. Our research is a proof of concept determining a new way of water classification combining optics concepts with machine learning model. This has been highlighted in the abstract, introduction and conclusion.

• A very weak analysis of the results obtained.

The authors would like to thank the reviewer for their valued comments. A paragraph dedicated for the results’ analysis has been added to the paper.

• The scientific nature of the article is doubtful. It deals with the concept of building a system for determining the salinity of water, which is closer to production than to science. There are no experiments confirming the calculations. There is no justification for the choice of sensor components. 

The authors would like to thank the reviewer for their valued comments. This research is a proof of concept of the proposed sensor design. It consists of a new way of water classification combining optics concepts with machine learning model. Different simulations using COMSOL have been performed, but no prototype has been developed at this time.

The obtained COMSOL simulation results have been validated by comparing them to values obtained from other experimental work and to analytical equations. Results showed a good agreement.

• There is no line numbering in the text of the manuscript.

The authors would like to thank the reviewer for their valued comments. Line numbering has been added.

• Questions and comments on the text of the manuscript:

1. There is no comparative analysis of water methods and sensors in the Introduction section. What is worse than other methods of water analysis? There is no discussion of other devices.

The authors would like to thank the reviewer for their valued comments. A section listing the different techniques employed for sensing pure and salty water has been added. The advantages and disadvantages have been presented as well. A comparison table has been added to highlight our research contribution and its advantage towards other techniques. 

2. The explanation in two paragraphs at the bottom of page 2 is elementary physics, corresponding to the level of a school textbook, not a scientific journal.

The authors would like to thank the reviewer for their valued comments. We think that the physics part addressed by the reviewer is a core concept used in this research. If it is not recommended to have it, we have no problem removing it from the manuscript.

3. Figure 2 needs correction: reduce the range on the ordinate axis, and on the abscissa axis you need to present a uniform scale, not a histogram.

The authors would like to thank the reviewer for their valued remark. Figure 2 axis has been updated.

4. The criteria for choosing the size of domain1 and domain 2 are not specified.

The authors would like to thank the reviewer for their valued feedback. We assumed that the minimum and maximum size of domain 2 is 5 cm and 10 cm, respectively. Therefore, we have performed different simulations covering the range from 5 cm up to 10 cm, with a minimum step of 1cm. Further criteria might be considered in a future work.

5. Figure 3 is also presented incorrectly along the abscissa axis.

The authors would like to thank the reviewer for their valued remark. Figure 3 axis has been updated.

6. Some paragraphs represent well-known information (for example, the last paragraph of page 6)

The authors would like to thank the reviewer for their valued feedback. This paragraph could give a general idea about the machine learning model for readers not in this field. If it is not recommended to have it, we have no problem removing it from the manuscript.

7. The criterion for determining whether the water is clean or salty is not specified if it was modeled only for S=0 and S=35.

The authors would like to thank the reviewer for their valued comments. Indeed, the clean water and salty water are modelled for S=0 and S=35 respectively. More details could be found in tables 3 and 5.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

As it was a major review and author has addressed all my concerns adequately. Author has improved the paper in structure wise, typo mistakes, grammatical errors and all other suggested changes. Still I have conflicts, reviewer is not able to understand the literature review section of the manuscript. As per author comments “Pure and sea salt water detection techniques” is a literature review section then make its heading as a literature review. Article is acceptable after minor changes.

Author Response

Response to Review

 

The authors would like to take this opportunity to thank the reviewer for their time, consideration and insightful feedback. All the reviewer’s comments and suggestions have been taken into consideration and addressed in the revised version of the paper. All changes have been highlighted in Green in the revised manuscript.

 

Reviewer#1:

• As it was a major review and author has addressed all my concerns adequately. Author has improved the paper in structure wise, typo mistakes, grammatical errors and all other suggested changes. Still I have conflicts, reviewer is not able to understand the literature review section of the manuscript. As per author comments “Pure and sea salt water detection techniques” is a literature review section then make its heading as a literature review. Article is acceptable after minor changes.

The authors would like to thank the reviewer for their valued comments. The title of the second section of the paper has been changed from "Pure and sea salt water detection techniques" to "Literature review", as recommended.

 

Author Response File: Author Response.docx

Reviewer 2 Report

Despite the authors' response, Figures 2 and 3 have not been corrected along the abscissa axes.

The scientific value of the article is still in doubt.

Author Response

Response to Review

 

The authors would like to take this opportunity to thank the reviewer for their time, consideration and insightful feedback. All the reviewer’s comments and suggestions have been taken into consideration and addressed in the revised version of the paper. All changes have been highlighted in green in the revised manuscript.

 

Reviewer#2:

• Despite the authors' response, Figures 2 and 3 have not been corrected along the abscissa axes.

The authors would like to thank the reviewer for their valued comments. Figures 2 and 3 abscissa axes have been changed from a histogram format, to a uniform scale.

 

• The scientific value of the article is still in doubt.

The authors would like to thank the reviewer for their valued feedback.

Here is a clarification of the scientific contribution of our research:

Most of the existing water salinity measurement techniques are either low-cost but not perfectly accurate, and could result in changing the water properties (chemical methods), or accurate but has a complicated design with a high cost (indirect methods: conductivity, density, SPR, optical, etc.).

Few studies proposed salt and pure water sensing techniques combined with Machine Learning, and were proven to be more accurate at a lower cost, without changing the solution’s chemical properties. However, using ML requires data preparation and scaling, which is a challenge for most of the proposed techniques.

Our proposed solution uses COMSOL to generate the data. Then, the data is validated using an analytical method. Consequently, the data is scaled and rearranged in way that the machine learning can find the weights (patterns and fingerprints) to classify the sample.

We are not aware of any existing research that applied the proposed combinations/steps.

 

To conclude, the proposed method has the following advantages:

- Biochemical-free technology, so chemical and electrical properties of the sample stay unchanged.

- Rich data obtained from the COMSOL simulation, which can be easily restructured for the different classification purposes.

- The sensor design concept is simple and provides high accuracy.

 

Author Response File: Author Response.docx

Round 3

Reviewer 2 Report

Figure 2 has been corrected. The abscissa axis of Figure 3 is still not represented correctly.

Author Response

Response to Review

 

The authors would like to take this opportunity to thank the reviewer for their time, consideration and insightful feedback. All the reviewer’s comments and suggestions have been taken into consideration and addressed in the revised version of the paper. All changes have been highlighted in green in the revised manuscript.

 

Reviewer#2:

• Figure 2 has been corrected. The abscissa axis of Figure 3 is still not represented correctly.

The authors would like to thank the reviewer for their valued comments. Figures 3 abscissa axis has been changed to a uniform scale, as per the reviewer recommendation.

 

Author Response File: Author Response.docx