Multi-Objective Optimization for Sustainable Pavement Maintenance Decision Making by Integrating Pavement Image Segmentation and TOPSIS Methods
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis paper is focus on pavement maintenance decision making based on CNN detection result and maintenance cost strategy. Unet CNN model was adopted to obtain the true area of defeat areas, while TOPSIS method was used to get the maintenance cost. But there is no information on the the detail maintenance strategy on defeat area. Will you use same maintenance strategy for all cracks? Is there any different between large and small crack area? Besides, dataset with limit 50 images is not enough for training high precision CNN network. So, I think the authors should improve their manuscript.
Comments for improving.
(1) In abstract, the author states that “Nevertheless, the existing automatic detection can only recognize and classify pavement distresses. However, they are unable to accurately determine the precise dimensions of specific distresses such as cracks and potholes”, please verify.
(2) In abstract lines 26-27,"This maintenance strategy achieved a substantial reduction of 30.80% in carbon emissions and a cost reduction of 20.81% compared to the highest values among all maintenance strategies." Please give your detail carbon and cost calculation method.
(3) In section 3.1, the dataset with 50 images is not enough. Besides, state your method on how to calibrate your camera, it is very import for you to get area of crack or pothole.
(4) In section 4.1, will you get overfit CNN result with 500 epochs in small dataset? Verify your performance and your training parameters.
(5) Taking crack as example, what are the difference among crack maintenance strategies? Will the Unet detection result affect the strategy?
(6) In Fig.17, verify the relationship between the cost and carbon emission, it is far beyond my math knowledge.
Author Response
Dear reviewer,
Thank you for your comments and suggestions, please see the attachment for our response.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe study on the "Multi-objective optimization for pavement maintenance decision-making by integrating pavement image segmentation and TOPSIS methods" is valuable and offers an engaging challenge. The following suggestions were offered to enhance the quality of the paper after a thorough review.
- There are mistakes in the format of the paragraphs that need to be modified, such as line 24, 112.
- Line 26, the unit of score should be added.
- The context of speech in lines 36, 37 is general, so there is no need to cite the reference.
- The article's structure should be presented towards the end of the introduction. Add a paragraph at the end of the introduction that includes 1-gaps in the background that you are attempting to fill, and 2-your novelty and distinctive features 3- the reasons of TOPSIS using should be mentioned.
- The details of U-Net algorithm and TOPSIS method should be added.
- Line 182, the weight of the distress should be clarified further, , so readers should be able to understand what the authors mean.
- In figure 4, the image of sealing should be replaced because it not clear.
- There are mistakes of all equation captions, should be modified.
- 50 images are few to reflect the data of the case study for applying the research methodology.
- Line 422, the dimension photographed picture is small, it should contain at least the width of the lane.
- Model training involves an accurate setting of several hyperparameters, namely batch dimension, epochs, and so on. The Method of selecting the best hyperparameters should be added.
- In figure 13, test curve should be added.
- Figures 16,17 are incorrectly formatted.
Comments for author File: Comments.pdf
Author Response
Dear reviewer,
Thank you for your comments and suggestions, please see the attachment for our response.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThis study endeavors to develop a multi-objective decision-making method for pavement maintenance based on the pavement image segmentation and TOPSIS methods. Pavement condition, carbon emissions, and costs are taken into consideration in the multi-objective model. I think it is a good idea to use pixel area to determine the actual dimensions of pavement distress, which allows for lightweight and rapid automatic pavement distresses detection. But there are still the following comments that need to be addressed.
1. The authors mentioned 108 possible maintenance treatments in the last part of section 2.2, but only gave a table of maintenance strategy selection in Table 2. The authors should explain that how these specific possible maintenance treatments are combined?
2. Table 7 indicates the energy consumption of machinery, normally, a loaded vehicle is also a type of machinery, why is there no energy consumption of loaded vehicles in Table 7?
3. There are more types of pavement damage than just potholes and cracks, is it reasonable to use only them to reflect the pavement condition index?
4. The authors only calculate the carbon emissions of various maintenance strategies, it is suggested to use “carbon emissions” instead of environmental impacts.
5. It is recommended to add theoretical knowledge of pavement maintenance engineering in the Introduction, including What does pavement maintenance engineering refer to? What is the importance of maintenance engineering in terms of automatic distress detection, environmental impacts, and economic impacts?
Comments on the Quality of English Language
ok for me.
Author Response
Dear reviewer,
Thank you for your comments and suggestions, please see the attachment for our response.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe reviewer appreciates the corrections made by the authors. However, modifications are still required. Camera is very important for the authors’ result; however, they didn’t consider to calibrate the camera before using it to detect crack or potholes’ actual size. So, add camera calibration method.
Author Response
Dear reviewer,
Thank you for your comments and suggestions, please see the attachment for our response.
Author Response File: Author Response.pdf
