Modeling the Cause-and-Effect Relationships between the Causes of Damage and External Indicators of RC Elements Using ML Tools

Round 1

Reviewer 1 Report

This paper focuses on the application of machine learning (ML) tools to identify the cause-and-effect relationships between the external indicators of reinforced concrete (RC) structures and their damage causes. The authors compared the performance of four ML models, namely Support Vector Regression (SVR), Decision Trees (DT), Random Forest (RF), and Artificial Neural Networks (ANN) and found that ANN had the best results. The authors optimized six ANN models and found that the ANN model with ADAM loss function and sigmoidal activation had the best performance with MAPE 3.38% and R2 0.969. The authors emphasized the novelty of their study as it developed an ML model based on ANNs, which can diagnose the technical condition of RC elements, helping in timely maintenance and repair.

The paper provides an interesting and valuable contribution to the field of identifying and diagnosing defects in RC structures using ML models. The authors have clearly described the methodology and results. However, there are a few points that could be improved:

·         The introduction section can be strengthened by clearly identifying the research hypothesis, research questions, objectives, motivations, and novelty of the work. It will help the reader to understand the context and purpose of the study better.

·         The paper can be made more visually appealing by adding a flowchart or a graphical abstract that summarizes the methodology and findings of the study. It will help the reader to comprehend the study quickly.

·         The authors have provided a detailed explanation of the results obtained from the ANN model, but the paper lacks a discussion on the implications of these results. The authors can elaborate on the practical implications of their findings and how they can be useful for the industry.

 

·         The authors have claimed that their study is novel, but they have not provided any evidence to support this claim. The authors can elaborate on how their study is different from previous studies and how it contributes to the field.

Author Response

Dear Reviewer,

We sincerely thank you for your time and useful suggestions to improve our research. We tried to take into account all your comments and remarks.

with best regards Authors

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper is generally well organized and written with some spelling issues that can be solved easily by an in-depth re-reading. Therefore, the paper only needs a minor revision according to the following comments

 

The abstract only gives an overview of ML techniques investigated with no details about the structures investigated to evaluate their performances. Some brief detail would be appreciated.

Lines 114-117: design flaws are among the causes too

Line 180: what is RK?

Section 2.2. devoted to describing the cause-and-effect relationships between external indicators of damage and their causes presents a very in-depth analysis although it neglects the internal material properties’ variability of structures. To this end, the authors are suggested to refer to and possibly cite the following article (10.3390/ma12121985) to demonstrate the effects of within-structure variability.  

Line 104: “Currently, there are a different Machine Learning” should be “Currently, there are different Machine Learning…”

It seems that some issues are related to the need of an English revision. For example on line 233 “The scientists have used various ML” Should be “Scientists have used various ML”. please check the whole Manuscript.

Line 394 and 397: “corrosion of reinforce” shall be “Corrosion of reinforcement”

 

Conclusions: discussing in the last part the effectiveness of ML methods for identifying the cause and effect relationships should not be regardless of visual inspections of skilled engineers as well as of permanent monitoring, especially for high-importance and strategic structures.

Author Response

Dear Reviewer,

We sincerely thank you for your time and useful suggestions to improve our research. We tried to take into account all your comments and remarks.

with best regards Authors

Author Response File: Author Response.pdf

Reviewer 3 Report

Please find the following points;

 

1-      The paper is very well written , organizes, include with the attractive highlights.

2-      The abstract is looking weak owing to the unavailability of background knowledge of this study. Add comprehensive line at the start of the abstract about the background history of the work. Also, add some key values from results and highlight the novelty of this work clearly. The ending of this section is quite abrupt. Complete the abstract with a conclusive on this work and its findings.

3-      The introduction needs to be more emphasized on the research work with a detailed explanation of the whole process considering past, present and future scope. How the present study gives more accurate results than previous studies? It needs to be strengthened in terms of recent research in this area with possible research gaps. It is strongly recommended to add a recent literature.

4-      Introduction gives some context but does not motivate the research problem well. As a result, the paper reads less convincing and compelling; try to add some research paper to your manuscript.

5-      Research gaps should be highlighted more clearly and future applications of this study should be added.

6-      Please discuss about more the database, especially the statistical analysis of the parameters. Did the author think about the non-linear correlation between the selected parameter?

7-      Please discuss about the limitation of the database, its impact on the training of the prediction models and the method for the validation of the prediction tool. Also how these limitations will affect the accuracy of the propose model.

8-      Please avoid the basic details about the methodology in the introduction portion, the introduction portion, please use only the latest reference. Please reduce these sections.

9-      The conclusion should be an objective summary of the most important findings in response to the specific research question or hypothesis. A good conclusion states the principle topic, key arguments and counterpoint, and might suggest future research. It is important to understand the methodological robustness of your study design and report your findings accordingly. Please improve your conclusion section.

10-  Please re-make Figure 3 for better presentation? Please increase the size of the font of figures 1,2 and 3. Please improve the quality of figures and use ONE software for all figures.

11-  Please include some graphical explanation of the comparison section.

 

The author needs to address the abovementioned points for the betterment of the manuscript. 

Author Response

Dear Reviewer,

We sincerely thank you for your time and useful suggestions to improve our research. We tried to take into account all your comments and remarks.

with best regards Authors

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

The authors responded well to the suggested comments; please accept the manuscript.