Development of Motorway Horizontal Alignment Databases for Accurate Accident Prediction Models
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe paper applied a flow of procedures working on a Portugal highway to 1) extract a refined horizontal curve database, then compared with the existing curve database, and 2) developed safety performance functions, and analyzed the propagation of error of independent variables. The paper is organized well in general but has several main issues. The motivation of the paper is unclear, thus why the authors develop a new curve database needs elaboration. From Section 4 Result and discussion, the authors had an existing inventory with horizontal curves in hand, but they chose to create their own database from scratch: starting from sketching centerlines from satellite imagery and going through a lot of manual procedures. The authors should explain their motivation for this approach: why they chose to create a new database rather than reusing/refining an existing one. The methodology section needs better clarification, especially how the authors conduct the error propagation. I recommend “Reconsider after major revisions” as the main idea looks promising but substantial systematic improvements are needed before publication.
I have the following comments:
Page 1 Line 39: “The high predictive power of …” This statement needs reference.
Page 1 Line 44: “However, in many cases, …” The inaccuracy issue may be commonly known but it would be helpful if the authors could add some examples/references to strengthen the point.
Page 2 Line 51-54: “Although a multitude of …” This is the research gap which motivates the paper, and it may need better clarification. For example, definition of “the effect of the error”, “accident prediction models” meaning the safety performance functions or the crash modification factors, how to estimate the reliability of prediction models, how to evaluate the accuracy of variables, such detailed information may need better elaboration.
Page 2-3 Background section: This section is generally fine. I wonder if the authors could include a discussion on how parameter estimation errors might occur and their potential impacts. For example, if a spiral transition curve is mistakenly treated as a circular arc—an error that can be common in curve detection—how might this affect the estimated radius and other parameters?
Page 3-4 Section 2.1 Alignment recreation: This section is generally fine with several data sources for curve recreation. However, it may be helpful to add a summary paragraph at the end, potentially with the pros and cons from each source to create curves.
Page 5 Line 176: Not sure about the “speed prediction models”. The authors may need to clarify.
Page 4-5 Section 2.2 Geometric design and safety: I can feel a disconnection in this section. The discussion on variables commonly used in safety performance functions was good. However, the last paragraph (Page 5, Lines 174-181) seems out of place, and it’s unclear what the authors are trying to convey there.
Page 6 Figure 2: It would be helpful if the authors specified which steps used the existing data sources and which steps were created by the authors themselves.
Page 10 Line 341: What is the point mentioned the roadway diverge and rejoin again, and using Figure 7 to illustrate? If the roadways are separated, the curves should be individually located on both sides of the roadways, right?
Page 12 Equation 10, Line 390-393: Not sure how this equation came up. The authors may need better clarifications. “The error value is -5.6% …” The authors may need to use plain language to explain what the statement means.
Page 13 Figure 8: How to read these two figures? What do the numbers and the lines represent? The whole section of how to examine the relative error propagated needs better clarification in the methodology section.
Author Response
Please, see the attached file.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe manuscript presents an innovative and well-written methodology for obtaining accurate horizontal geometric features from digital imagery, which has been verified through rigorous numerical and statistical procedures. However, minor revisions are needed before publication.
It is essential to enhance the paper by including more recent references.
I would like to include the aspect of road design, in the introductory part, e.g.
In the initial phase of road design (Simić et al., 2023), it is crucial to prioritize traffic safety (Stević, et al, 2022; Trifunović et al., 2024), above all other considerations, while also paying careful attention to traffic flow dynamics (Jayapal et al., 2023; Noorlaelasari et al., 2023) and accommodating the increasing prevalence of electric vehicles (Puška et al., 2023) to ensure a sustainable and efficient transportation network.
Noorlaelasari, Y., Juarti, E. R., Rahman, A., & Juraida, A. (2023). Assessing Compliance to Minimum Service Standards of Indonesian Toll Roads: An Examination of the Padaleunyi Toll Road. J. Urban Dev. Manag., 2(2), 84-94. https://doi.org/10.56578/judm020203
Puška, A., Stojanović, I., & Štilić, A. (2023). The Influence of Objective Weight Determination Methods on Electric Vehicle Selection in Urban Logistics. J. Intell. Manag. Decis., 2(3), 117-129. https://doi.org/10.56578/jimd020302
Simić N, Ivanišević N, Nedeljković Đ, Senić A, Stojadinović Z, Ivanović M. Early Highway Construction Cost Estimation: Selection of Key Cost Drivers. Sustainability. 2023; 15(6):5584. https://doi.org/10.3390/su15065584
Trifunović, A., Senić, A., Čičević, S., Ivanišević, T., Vukšić, V., & Simović, S. (2024). Evaluating the Road Environment Through the Lens of Professional Drivers: A Traffic Safety Perspective. Mechatron. Intell Transp. Syst., 3(1), 31-38. https://doi.org/10.56578/mits030103
Stević, Ž., Subotić, M., Softić, E., & Božić, B. (2022). Multi-Criteria Decision-Making Model for Evaluating Safety of Road Sections. J. Intell. Manag. Decis., 1(2), 78-87. https://doi.org/10.56578/jimd010201
Jayapal, P. K., Muvva, V. R., & Desanamukula, V. S. (2023). Stacked Extreme Learning Machine with Horse Herd Optimization: A Methodology for Traffic Sign Recognition in Advanced Driver Assistance Systems. Mechatron. Intell Transp. Syst., 2(3), 131-145. https://doi.org/10.56578/mits020302
I also suggest one more aspect to be improved in the introduction:
The prediction of traffic accidents, using new technologies, can have a great impact on the improvement of traffic safety (Kodepogu et al., 2023).
Kodepogu, K., Manjeti, V. B., & Siriki, A. B. (2023). Machine Learning for Road Accident Severity Prediction. Mechatron. Intell Transp. Syst., 2(4), 211-226. https://doi.org/10.56578/mits020403
To enhance clarity and organization, this chapter, "Materials and Methods," will be divided into several subsections. Additionally, an introductory section will be included to provide a comprehensive overview of the methodologies employed and the rationale behind their selection.
In Figure 2, please outline the section where the Street View inspection occurs with a rectangle. This will help to clearly define all three components, including both the entrance and exit, assuming my understanding of the image is accurate.
Praise for the results of the work.
Some figures and tables are out of format, check the formatting.
In the conclusion, write in more detail the application of the obtained results in practice, future research and the limitation of the study.
Author Response
Please, see the attached file.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for Authors This research proposes a methodology to obtain accurate horizontal geometric features from digital imagery based on the analysis of planimetry, feature geolocation and centerline azimuth sequence. The topic of this paper is very interesting and the manuscript is easy to follow. I have some comments. 1. The errors of the proposed method is not well presented. Perhaps RMSE or other merics should be added. 2. A comparison between the presented study and state-of-art methods is encouraged. 3. Most of the references are rather old. 4. A detailed discussion of the crash prediciton models is encouraged to show how the errors influence the crash rates in different sections.Author Response
Please, see the attached file.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThe authors have addressed all my questions. I don’t have further questions.
Comments on the Quality of English LanguageThe English is good.
Reviewer 3 Report
Comments and Suggestions for AuthorsAll my comments have been addressed
