Method for Select Best AIS Data in Prediction Vessel Movements and Route Estimation

Round 1

Reviewer 1 Report

In the present manuscript, the authors present a sample method to select historical data on vessel-specific routes to optimize the computational performance of the prediction of vessel positions and route estimation in real-time. The level of English is not adequate, so it is requested that it be corrected. However, it is a previously submitted article, for which several corrections have been made. Next, I will present some aspects to improve:
-Are the routes presented the most optimal? if they are the most optimal against which can they be compared and how optimal is it?
-Authors must write "Figure" in full, since it must be standardized, although writing it completely or in abbreviated form.
-The Figures must be large to be able to appreciate them correctly, especially those that show significant graphics to demonstrate what is presented by the authors. In addition, in these Figures they must write the units on the abscissa and ordinate axes.
-There are two enumerations of Tables that are repeated. It must be corrected.
-Table 2 is misplaced. This Table should be close when it is mentioned.
-Improve the conclusions.

Author Response

Dear Reviewer 1,

We confirm that this work is original and has not been published elsewhere, nor is it currently under consideration for publication elsewhere.

We want to submit a manuscript as a research article enclosing herewith a manuscript entitled "Method for Select Best AIS Data in Prediction Vessel Movements and Route Estimation."

Round1

• The level of English is not adequate

We review the English in the entire article.

 

• Are the routes presented the most optimal? if they are the most optimal against which can they be compared and how optimal is it?
-

 

Yes, there are the most optimal.

The routes are not compared. They are selected with the range of the target route.

The most significant variables for to perform the selection are trimmed, MAD, and range:

The trimmed mean is less sensitive to outliers. It allows to discard them and uses the distribution data to get closer to the target route.

MAD is less sensitive to outliers. It is an important variable to relate the routes that are statistically similar to the target route.

The range is susceptible to outliers. This property allows discriminating those routes that have large deviations outside the target route.    

.           Optimal refers to after the process of the DBSCAN, the method discards all routes that do not resemble the route to predict. The best routes (series of positions) have a similarity in statistical properties identified in the clustering process.

From line 282 to line 337 on the article is the description of the selection of the routes.

• Authors must write "Figure" in full, since it must be standardized, although writing it completely or in abbreviated form.

We correct it in abbreviated form.

 

• The Figures must be large to be able to appreciate them correctly, especially those that show significant graphics to demonstrate what is presented by the authors. In addition, in these Figures they must write the units on the abscissa and ordinate axes.


 

We scale the figures and the units on the abscissa and ordinate axes.

 

• There are two enumerations of Tables that are repeated. It must be corrected.
-Table 2 is misplaced. This Table should be close when it is mentioned.


We fixed the latex command.

• Improve the conclusions.

We improved the conclusions since line 466.

The results are from 80.5% to 84% in the accuracy of the prediction. The amount of information used was more than the first example. That allowed the accuracy to be better, at 84%. The historical AIS data samples are similar from the block to be estimated, out of 10 predictions. The absolute mean difference for latitude is 0.0015655 degrees, while the absolute mean difference for longitude is 0.00211949 degrees.

 

The results concern the assets of the vessels that move in the seas, especially in economic terms. The paper shows the PreMovEst method's application and testing of its accuracy using ARNN and MICE techniques. Daily vessels move billions of items across the oceans from one country to another, one of the drawbacks of handling all this information.

The process of data selecting to train the neural network was in function of the number of trips made by the vessel from a port of departure to the arrival port.

One of the challenges to be undertaken is to obtain better results concerning the accuracy of the execution time savings.

Future work is to accelerate the prediction model with Graphical Process Units (GPU). It also requires image processing, which implies supercomputers to process such data, limiting its effectiveness for real-time monitoring of vessels navigating within a designated maritime area—increasing the amount of processed information.

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper presents the method of application and testing of its accuracy via ARNN and MICE techniques. Overall, it is a qualified paper, I therefore recommend that this paper can be accepted.

 

Author Response

Dear Reviewer 2,

We confirm that this work is original and has not been published elsewhere, nor is it currently under consideration for publication elsewhere.

We want to submit a manuscript as a research article enclosing herewith a manuscript entitled "Method for Select Best AIS Data in Prediction Vessel Movements and Route Estimation."

Author Response File: Author Response.pdf

Reviewer 3 Report

This article is about AIS data and Artificial Neural Network. This work aims to present a prediction model that correctly predicts the physical movement in the route. It supports route planning to the Vessel Traffic Service. 

Figure 7 must be improved because it is difficult to determine positions of the points in 3D.

I recommend to extend the list of references. Look e.g. at https://doi.org/10.15837/ijccc.2020.3.3864

Author Response

Dear Reviewer 3,

We confirm that this work is original and has not been published elsewhere, nor is it currently under consideration for publication elsewhere.

We want to submit a manuscript as a research article enclosing herewith a manuscript entitled "Method for Select Best AIS Data in Prediction Vessel Movements and Route Estimation."

• Figure 7 must be improved because it is difficult to determine positions of the points in 3D.

We improved the Figure 7.

• I recommend to extend the list of references. Look e.g. at https://doi.org/10.15837/ijccc.2020.3.3864

We added the proposed reference and another one on lines 542 and 551.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thanks to Authors for performing the changes suggested by Reviewers. However, Figure 3 and 4 have to be bigger because for reader can be difficult for understands detaily.

Author Response

Dear Reviewer3,

We made Figures 3 and 4 bigger. 

The authors, 

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Peer Review Report

Ms. Ref. No.: applsci-1078499

Title: Inspiration Nature Method for Select Best AIS Data in Prediction Vessel Movements and Route Estimation

 

Authors: Rogelio Bautista-Sánchez, LilianaIbeth Barbosa-Santillan, Juan Jaime Sánchez-Escobar

The subject presented in the manuscript is very interesting. The subject of the article is within scope of the journal. However, in general, the authors should be more precise in my opinion – more descriptions should be included in this manuscript. Then, it can be submitted once again. Currently, there are too many figures in comparison to the text of the manuscript. Because of that, I recommend the paper for rejection. I believe that the authors will find below some suggestions, which will help them to improve their manuscript:

 

Major comments:

• Please avoid using abbreviation in the titles of the paper.
• Introduction should be extended and should be based on more literature references. In my opinion at least 30 papers/books should be here cited.
• Why the method is called “Inspiration nature method” by the authors? I suggest to change the name, because it suggest it has something to do with “nature inspired methods/algorithms”. Therefore, in my opinion, it is very misleading name of the method.
• The authors didn’t write anything about implementation? How the authors implanted the method? Used any commercial software or writing any own programs (in which programming language)?
• In my opinion, this paper should be based on more equations (any equations)?
• The authors should described more deeply in the text what include Tab. 1.
• 3 and Fig. 4. are unreadable. More description in text is needed.
• 6 is cut out.

Minor comments:

• Line 83 – “chained” with a capital letter.
• Line 98 – after “In this way, …” it should be “it is ” instead of “is”.
• Line 109 – bold.
• Line 144 – dot instead of comma.

Conclusion:

The subject of the paper and the manuscript are very interesting. However, the manner of presentation of the subject is in my opinion not enough to submit to scientific journal in its current form. The manuscript needs to many improvements. Because of that, I recommend the manuscript for rejection.

Author Response

Dear Reviewer,

We attached the highlighted article.

Best Regards,

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Remarks are placed at the higlighted text in attached manuscript.

Generally, too much general words that do not explain what they mean and what are You talking about.

You must add definitions of all specific words, in between description of weights, scales of parameters and many others.

Most serious:

1. I cannot trace where in the text is described question of "computational infrastructure savings" and the gains of the presented work/methods used for it.

2. The results of predictions are moving through the land. It disqualifiy the method from being put into practice.

So what it is to serve this work (article)?

 

 

Comments for author File: Comments.pdf

Author Response

Dear Reviewer,

We attached the highlighted article.

Best Regards,

Author Response File: Author Response.pdf

Reviewer 3 Report

In the present manuscript, the authors present a sample method to select historical data on vessel-specific routes to optimize the computational performance of the prediction of vessel positions and route estimation in real time. However, I will comment on some aspects to improve the article:
-The authors are not using the format of the journal correctly, since the logo of the journal to which they are sending or the name of the journal does not appear in the footer. Also in some parts the subsections are in bold (what is with index i, ii, iii, iv).
-The authors are incorrectly writing the acronyms in the document. The correct spelling of the acronym is the first capital letter of the meaning of the acronym, for example, "Vessel Traffic Monitoring and Information Systems (VTMISs)". All acronyms in the manuscript must be corrected.
- There are acronyms that have no meaning, such as "COVID", among others. Authors should provide their meaning for a better understanding of the reader.
-The Introduction Section should be improved, and include a Related Works Section. Apparently, the authors have combined it into a single section, so it should be separated.
-Authors must change "Part" to "Section" in line 72.
-Authors must write with capital letters when referring to Section, Algorithm, Table, Figure, Equation, and completely within a scientific manuscript.
-It is not correct to place line 90 and 91, which details a URL, for that they are necessary to cite them.
-Figure 3, and 4 are horrible to understand, the images are extremely small.
-The title of Figure 5, 6, should be a title that explains in very few words, not a 3-line explanation. For the explanation there is the text of the manuscript.
-Figure 6, 7, 8, are crop from the image, and the legend must be in ascending order of the routes. Also, the axis units are required.
-What type of clustering are the authors using? They must give a brief explanation.
-The authors must demonstrate with more examples the prediction of the generated route.
-The conclusions are too small, they must be improved.
-Future works must be in the Conclusions Section, at the end of the conclusions presented by the authors.
-The authors must increase the references, they are very few.

Author Response

Dear Reviewer,

We attached the highlighted article.

Best Regards,

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

I'm still not convinced that the paper show appropriate level of scientific literature background description and methods proposed and applied in the paper. Because of that, I recommend to reject the paper in its present form.

Author Response

Dear Reviewer,

We have rewritten related works as having suggested and highlighted the corrections on the A_Reviwer1 pdf file.

Best Regards,

Author Response File: Author Response.pdf

Reviewer 2 Report

Few remarks I have placed in Your manuscript file.

Comments for author File: Comments.pdf

Author Response

Dear Reviewer,

We highlighted the corrections on the A_Reviwer2 pdf file.

Best Regards,

Author Response File: Author Response.pdf

Reviewer 3 Report

Thanks to the authors for performing the relevant changes. However, in the References Section, they must be in the corresponding MDPI format. Furthermore, the authors did not present a Cover Letter as it must be for the reviewers, they only sent the same manuscript highlighting the changes, but not giving a justification in many questions.

Author Response

Dear Reviewer,

We add the cover letter as you required.

Best Regards,

Author Response File: Author Response.pdf