Smartphone-Based Assessment of Bicycle Pavement Conditions Using the Bicycle Road Roughness Index and Faulting Impact Index for Sustainable Urban Mobility

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have conducted a detailed study of the assessment of bicycle pavement conditions using smartphones and two indices. The author has done a good job in covering most of the aspects. I have below comments for the authors.

1. Is R intentionally capitalized in “Risky conditions” inthe  abstract?
2. Line 52-53, there are research conducted which has adopted IRI for bicycle pavement using smartphone, for example https://doi.org/10.3390/s18030914 (number 7 in your reference list). So I do not agree with the statement that it is ill-suited.
3. Line 60-62 There are a lot of studies that also assess the perception of bicycle comfort levels. For example https://doi.org/10.3390/su15054495 and https://doi.org/10.1016/j.jclepro.2018.04.275
4. Line 62-63 which majority studies, the authors have not given us details of any such studies in the introduction.
5. The limitation suggested by the authors (as I have given two references) has been addressed. I would suggest providing a solid reasoning/justification for this study.
6. Not enough literature is available to the reader on the topic.
7. From line 81-65 discuss the methodology and should not be in the introduction.
8. Where is section 2 as the authors directly started from subsection 2.1.
9. Please clarify whether the test sites were selected for experimental or to capture variability, as the current rationale presents conflicting objectives. Line 103-105 and 110-112.
10. Can the authors map selected bicycle routes, showing the entire segment of the selected path, not just the location? The current maps are not helpful.
11. I only see two types of pavement used: CP and AP. While there exist other types of bicycle paved paths/routes, such as paving tiles, this is a critical oversight.
12. Have to be sure if there was surface damage as mentioned in Table 1. A figure with images would be useful.
13. Also, check this recent study (https://doi.org/10.3390/s24227210), which has used the Physics Toolbox Sensor Suite, to justify why this specific app was used.
14. Why were 40 cyclists used for the panel survey? Why not more?
15. The scale used is wrong. For the question perceived surface smoothness, it should be either from Very good - very bad or Very comfortable Very – Uncomfortable
16. Use high quality picture the Figure 3 is not visible.
17. The line 312-317 are redundant as previously explained in methodology section. The last line also belong to methodology section.
18. Line 3469-347: What does the strong inverse correlation imply? Need explanation so that readers can understand the inverse relationship.
19. In my opinion, section 3.4 is redundant; these points are already explained above.
20. I missed a critical discussion of the results in the discussion section. It reiterates the results again.
21. The conclusion also reiterates the results. This makes the manuscript repetitive. Please rewrite it again, only focusing on what you conclude from your study. Only pick the main result (not all, please).

Author Response

Thank you very much for your thoughtful and detailed feedback. Your comments significantly helped us enhance the structure, clarity, and scholarly depth of the manuscript—especially in the literature review and justification of our methodological framework. We have carefully addressed each point and revised the manuscript accordingly.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The research methodology is well-designed and appropriate for the research questions, demonstrating a solid understanding of the subject area. The data analysis is thorough and clearly presented, with relevant statistical tests applied. The conclusions are supported by the results and offer valuable insights. However, I suggest the author include more detailed descriptions of the experimental materials and procedures to improve reproducibility. Additionally, discussing potential limitations and avenues for future research would strengthen the manuscript further.

Author Response

Thank you very much for your thoughtful and detailed feedback. Your comments significantly helped us enhance the structure, clarity, and scholarly depth of the manuscript—especially in the literature review and justification of our methodological framework. We have carefully addressed each point and revised the manuscript accordingly.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This article presents a smartphone-based framework for evaluating bicycle pavement conditions using two indices: the Bicycle Road Roughness Index (BRI) for ride comfort and the Faulting Impact Index (FII) for shock severity. Data were collected through tri-axial accelerometers and validated against cyclist perception via a structured panel survey involving 40 riders. Regression models linked objective measurements to subjective ratings, leading to the development of five risk and rideability levels. Field tests across multiple sites in Korea confirmed the method’s accuracy and practical value. The framework enables scalable, low-cost pavement monitoring and supports data-driven maintenance planning to improve cyclist safety and urban mobility.

1. Section 2 title “2.1 Study Area” appears before “2” is introduced. Please fix hierarchy formatting.
2. The manuscript jumps between methods and results in several places (e.g., pg. 5). Please clearly separate the two.
3. A dedicated Methodology section header (Section 2) would help frame subsections 2.1–2.5.
4. Another dedicated section of the Literature review is a must.
5. Tables are not consistently placed or referred to. Please ensure they appear close to the discussion and are cited.
6. The Discussion section repeats several points from the Results. Please restructure for new insights only.
7. Clarify “step difference” in layman’s terms when first introduced.
8. The phrase “perceptual breakpoint” (pg. 15) is vague. Please define.
9. Use consistent terms for rating scales: “Likert scale” vs. “panel rating” vs. “subjective score.”
10. Use consistent notation for units. e.g., always “m/s²” instead of “m/s2” or “m/s² ” (with or without space).
11. Explain why 50 repetitions per fault height were chosen. Justification is needed.
12. FII values jump significantly. Please explain if any smoothing or outlier correction was applied.
13. In Table 5, check for consistency: FII for 20 mm at 20 km/h appears lower than 15 km/h. Double-check values.
14. The threshold of “87.3 m/s²” for FII seems very precise. Please explain its derivation or round for practicality.

 

Author Response

Thank you very much for your thoughtful and detailed feedback. Your comments significantly helped us enhance the structure, clarity, and scholarly depth of the manuscript—especially in the literature review and justification of our methodological framework. We have carefully addressed each point and revised the manuscript accordingly.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript presents an innovative and practically relevant methodology for assessing bicycle pavement conditions using smartphone-based sensing combined with user perception. The dual-index approach (BRI and FII) is well-structured and offers useful implications for infrastructure monitoring and maintenance planning. The empirical design and data collection process are generally rigorous and clearly described. However, there are several aspects that should be improved or expanded upon prior to publication:

1. While the paper offers a solid methodological contribution, the literature review remains somewhat narrow and lacks engagement with the broader body of research on cycling infrastructure assessment, safety, and urban mobility. Currently, the discussion is primarily focused on sensor methods and roughness evaluation. For a journal like Sustainability, the study would benefit from a more comprehensive contextualization of cycling infrastructure within the wider field of sustainable transport and public health. In particular, I recommend integrating studies that examine the systemic role of cycling in multimodal transport networks and public health. Relevant works such as Bike-Sharing as an Element of Integrated Urban Transport System, and Health Impacts of Bike-Sharing Systems in the U.S. provide valuable perspectives on how cycling—especially in shared-use systems—interacts with broader urban and transport systems. These references would help frame the practical importance of infrastructure quality in real-world planning and policy contexts.

2. While the smartphone-based sensing approach is promising and cost-effective, the paper would benefit from a more detailed discussion on potential variability across different smartphone models, mounting configurations, and hardware sensitivity (e.g., accelerometer sampling rates, sensor drift). Currently, the methodology assumes consistent data quality, but this may not hold in real-world deployments involving diverse devices or users.

   It may also be helpful to briefly reference how the setup could be standardized, or whether inter-device variability was tested during pilot studies. This would strengthen confidence in the generalizability of the framework for large-scale implementation.

3. Check for redundancies in phrasing (e.g., “empirical threshold derived empirically”). Review typographical consistency in subheadings (some switch between title case and sentence case).

By addressing these points, the manuscript would be significantly strengthened, both in terms of scholarly contribution and presentation.

Author Response

Thank you very much for your thoughtful and detailed feedback. Your comments significantly helped us enhance the structure, clarity, and scholarly depth of the manuscript—especially in the literature review and justification of our methodological framework. We have carefully addressed each point and revised the manuscript accordingly.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I appreciate the reviewers addressing the comments; the manuscript has been significantly  improved, and I have a few more comments for the reviewer.  

1. The changes made in response to my comment#2 in my previous review, “while IRI provides a standardized method for assessing longitudinal roughness, several studies have highlighted its limited sensitivity to short-range discontinuities at low cycling speeds, and its application often requires high-precision equipment or multi-sensor configurations [5–7]. As this study prioritizes simplicity and field practicality using only a smartphone and a bicycle mount, a tailored model was required to align with these operational constraints.” I do not see those changes in the revised manuscript.
2. Please always provide the line and page number when any changes are made in response to comments. Now it is very hard to track the changes in the manuscript.
3. Line 64-65 The authors claim that the study reference (8) has lacked sensitivity to short-range discontinuities such as uplifted joints, potholes. Check the study again the have their results showed that it was able to correctly recognize all of the potholes/humps. The authors are advised to carefully review to make claims in the paper.
4. Line 90, why is “Four Major Rivers” with capital first letters?
5. A North Arrow and scale are missing in the maps.
6. Line 92, which “minimal environmental interference” are authors talking about. Specify them.
7. In response to my comment about the pavement material, I appreciate the answer; however, please note this in your limitation these other types of paved bicycle paths were not available in the study area.
8. Line 156 and 159 Which customized sensing/alternative app was used? Authors are should add more information.
9. I see the pictures are still very poor now in figure 3. It is strongly advised to make a sketch of the whole concept bike in section 2.2, this will simplify the reproducibility of the research. Then leave these two pictures in the section.
10. I am still missing a critical discussion of the results/system with other studies. Are your results in line with previous studies, or are there any contrasting results? If yes, then why? If this system provides better results, what similar or different thing in the literature was done? Moreover, how your study improved these shortcomings this is what I meant by critical discussion.

Author Response

Dear Editor,

On behalf of my co-authors, I am pleased to submit our manuscript entitled
“Smartphone-Based Assessment of Bicycle Pavement Conditions Using the Bicycle Road Roughness Index and Faulting Impact Index for Sustainable Urban Mobility”
for consideration for publication in Sustainability.

In this manuscript, we present a novel, smartphone-based dual-index framework for evaluating the condition and safety of bicycle pavements. Our approach introduces two new indices—the Bicycle Road Roughness Index (BRI) and the Faulting Impact Index (FII)—that are both empirically calibrated against user perception. By leveraging only off-the-shelf smartphones, our system enables cyclist-centered assessment of both sustained ride comfort and acute vertical hazards in a practical and scalable manner. We demonstrate that our dual-index method provides significant improvements over traditional roughness measures, such as the International Roughness Index (IRI), particularly in its ability to detect short-range surface irregularities and to align evaluation outcomes with actual cyclist experiences.

Field validation was conducted on South Korea’s national bicycle network, and our results show a strong correlation between the proposed indices and user satisfaction. The high spatial resolution of our data collection (100-meter intervals) also enables the detection of localized hazards that are often missed by kilometer-scale analyses, supporting more precise and actionable maintenance planning.

We believe that our findings will be of interest to the journal’s readership, as they offer a user-oriented, data-driven approach for supporting safer and more sustainable urban mobility. To the best of our knowledge, this is the first study to integrate a perception-aligned, dual-index system with smartphone-based sensing for large-scale, real-world bicycle pavement assessment.

This manuscript has not been published elsewhere and is not under consideration by any other journal. All authors have approved the final version of the manuscript and any potential conflicts of interest have been disclosed.

Thank you very much for your time and consideration.
Please do not hesitate to contact me if you require any additional information.

Sincerely.

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

1. It is annoying not to track the changes made, even though the line and page numbers are provided; however, they do not correspond to the submitted file. If there are discrepancies, you can always paste the exact changes made in the paper and your response in the response document.
2. Please check if these studies have used the Physics Toolbox Sensor Suite? Please be careful about such referencing.

“While commercially available apps such as Physics Toolbox Sensor Suite (v1.2.8) have been used in previous research [10,15].”

3. Thanks for adding the picture of the bicycle. Take a picture from an angle where the digital speedometer and a rear-facing action camera are also visible. Which I believe was part of the experiment.
4. I am referring to the response document. You are referring to two studies “We acknowledge that prior studies using IRI and single-axis vibration indices (e.g., Gao et al., 2018; Zang et al., 2018) have contributed valuable methods for quantifying general pavement roughness and detecting features like potholes However, these earlier approaches often lacked sensitivity to short-range discontinuities—such as uplifted joints and localized faults—which are critical for cyclists at lower speeds, and generally did not incorporate perception-based classification frameworks”.

Gao et al., (2018) included the perception of cyclists. Please check the study again.

5. Include references in the discussion part, which are very important when comparing. Now, authors are making a comparison, but with whom? No information is provided.
6. “Consequently, the explanatory power of previous methods with respect to user perception was moderate (often R² < 0.6), potentially limiting their practical utility for cyclist-centered infrastructure management”.

Which studies are the authors talking about? No references are provided.

7. Please provide reference for studies that measured the roughness using these devices, and then mention the limitation as you did already. “Unlike many existing systems that require expensive equipment or complex multi-sensor configurations (e.g., laser profilometers or IMUs), our method enables scalable, low-cost deployment without sacrificing diagnostic accuracy.”

Author Response

Please see the attachment.

Author Response File: Author Response.pdf