Machine Learning Applications for Reliability Engineering: A Review

Round 1

Reviewer 1 Report

The authors of "Machine Learning Applications to Reliability Engineering: a Review" aimed to make it easier to understand the difference between traditional modelling techniques and ML techniques and how they can be used for reliability applications. At the same time, I find it difficult to understand how the task is solved and what the main conclusions were made. To be eligible for publication in Sustainability, the review needs significant revision. In particular:

1. All abbreviations and acronyms (used by the authors both in the text and figures) must be decrypted after the first mention. For example, the abbreviations RAMS and PHM are used for the first time in the Abstract and at the beginning of the Introduction (line 35), and the authors provide their decipherment as early as Section 2.2.

2. Figures captions should be enlarged because it is difficult to see the information presented.

3. The text sometimes lacks references to figures and tables (for example, Figure 2, Table 3).

4. It is necessary to eliminate grammatical errors like “… de80searning…” (line 149).

5. The structure of the review needs improvement, as it is difficult to understand the relationship between the purpose set by the authors, the method and the order of information presentation. It is also worth clarifying which practical tasks of Reliability Engineering can be solved with the help of computational intelligence methods.

6. Section 2.3.2 needs significant revision. In particular, it is necessary to improve Figures 4 and 5, along with their description.

7. What means the R2 coefficient should be described (line 189, Figure 6).

8. The review contains the Figures, which require additional explanations in some cases (for example, Fig. 10).

9. Section 4 contains a significant number of Figures. The necessity for this requires additional argumentation.

10. Tables 3 and 4 should be combined. At the same time, it is necessary to explain why Data Type and Algorithm (Table 4) are presented only for 17 out of 19 analyzed articles (Table 3).

11. It is necessary to finalize the Conclusion based on the goal set in the work and argue the conclusions made.

Author Response

 "Please see the attachment."

Author Response File: Author Response.docx

Reviewer 2 Report

This review article presents an overview of machine learning strategies and their applications. It consists of several subsections focused on: the literature analysis approach; the basics of mathematical modeling and machine learning concept; the historical development of artificial intelligence and its applications; the theoretical concepts of supervised, unsupervised, reinforcement and deep learning; and finally with machine learning protocols and application examples. 

I will stress out some minor remarks:

- the abbreviations RAMS (Reliability, Availability, Maintainability and Safety) and PHM (Prognosis and Health Management) need to be explained when firstly mentioned in the manuscript text;

- the same is valid for the abbreviation IBM;

- Figure 8: the authors should check once more the literature, LDA is considered a supervised classification approach for data treatment;

- Line 251-252: the authors mention machine learning and artificial intelligence as separate/different term, whereas at the beginning of the manuscript they mention that these mean the same thing;

- Line 296: k-NN instead of k-nn;

- In my opinion, figure 14 is redundant, especially because there is so many;

- In my opinion table 3 is redundant;

- Table 4 needs to be restructured taking into account the same variables which could be placed in the same row;

- Lines 452-454: Please, check the sentence.

- There are not enough references for covering a topic for review article.

In my opinion, the major issue is that this work represents a basic approach to machine leaning, explaining some concepts that are almost considered a common knowledge for everyone doing research or applying machine learning in practice. Therefore, I would suggest the authors to submit their work in a lower quality journal.

Author Response

 "Please see the attachment."

Author Response File: Author Response.docx

Reviewer 3 Report

Comments for author File: Comments.pdf

Author Response

 "Please see the attachment."

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors of "Machine Learning Applications to Reliability Engineering: a Review" thoroughly revised the manuscript. They took into account all my comments. Therefore, I believe that the review can be published in the presented form.

Author Response

As requested by Academic editor : 

I increased figure texts in figure: 1 and 5
I reorganized Figure 6 as requested and increased text size.
Figure 10 and 11 were modified to increased texts and resolution and avoid copyrights issue.
I improved resolution for Figure 12, and 16 and decided to remove figure 13;
I propose to remove Figure 13 Feature extraction methods classification, because not much description is given, and it does not add much to the articles. Also, it would necessitate permission to use the copyrights.
For table 3,I added authors and year in a short format.
(Obtain a proper copyright release for Figures obtained from other references.):
I removed figure 13 and adapted figure 10 and 11 to avoid copyrights issue.
2. :
I modified the reference section accordingly and added the information as requested.

Reviewer 2 Report

The authors corrected the manuscript accordingly.

Author Response

As requested by Academic editor : 

I increased figure texts in figure: 1 and 5
I reorganized Figure 6 as requested and increased text size.
Figure 10 and 11 were modified to increased texts and resolution and avoid copyrights issue.
I improved resolution for Figure 12, and 16 and decided to remove figure 13;
I propose to remove Figure 13 Feature extraction methods classification, because not much description is given, and it does not add much to the articles. Also, it would necessitate permission to use the copyrights.
For table 3,I added authors and year in a short format.
(Obtain a proper copyright release for Figures obtained from other references.):
I removed figure 13 and adapted figure 10 and 11 to avoid copyrights issue.
2. :
I modified the reference section accordingly and added the information as requested.