Route Planning for Flexible Bus Services in Regional Cities and Rural Areas: Combining User Preferences with Spatial Analysis

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

（1）The manuscript effectively highlights the need for flexible and on-demand transport as a sustainable alternative in car-centric societies, particularly in rural areas where traditional public transport options are limited. However, the research objectives could benefit from further clarification in the abstract and introduction sections. Specifically, it would be helpful to succinctly define "flexible public transport" and the unique aspects of the "unexplored" context referenced, as well as to emphasize the importance of this study for future transportation planning in similar settings.

（2）The study employs a robust combinatorial approach, integrating quantitative social research and spatial analysis to gather both behavioral insights and spatial data relevant for route planning. While this approach is commendable, the methodology section could benefit from additional detail on the sample selection for the questionnaire survey, as well as on the spatial analysis techniques applied. Including more specifics on these aspects will enhance the transparency and reproducibility of the study's methodology.

（3）The analysis provides valuable insights into user preferences for flexible transport services, notably the preference for door-to-door service in the morning and stop-based service later in the day. This finding is insightful, yet further discussion on why this preference may exist (e.g., cultural or regional influences, work/school schedules) could deepen the reader's understanding. Additionally, while cost and time are highlighted as significant factors, more granularity on these influences (e.g., cost sensitivity across different income levels) could provide valuable context for policymakers.

（4）The study introduces a novel approach to route planning by combining spatial analysis with multicriteria decision-making tools, an advancement that can benefit flexible transport planning. Nonetheless, the manuscript would be strengthened by offering a comparative analysis with other route-planning methodologies in similar contexts, highlighting the advantages and potential limitations of the proposed approach. This would position the study more clearly within the existing body of flexible transport literature and underscore its unique contributions.

Overall, the paper presents a well-structured and innovative study that contributes valuable insights to the field of flexible transportation planning. Addressing the comments above will enhance the clarity, depth, and applicability of the findings, ultimately making the study more impactful for both researchers and transportation planners.

Comments on the Quality of English Language

Just some small polishing is enough

Author Response

(1) The manuscript effectively highlights the need for flexible and on-demand transport as a sustainable alternative in car-centric societies, particularly in rural areas where traditional public transport options are limited. However, the research objectives could benefit from further clarification in the abstract and introduction sections. Specifically, it would be helpful to succinctly define "flexible public transport" and the unique aspects of the "unexplored" context referenced, as well as to emphasize the importance of this study for future transportation planning in similar settings.

Thank you for the time you dedicated to reviewing our article. We have uploaded an improved version of the manuscript with all the changes marked.

Regarding the above comments, we tried to address it by modifying the first four sentences of abstract. In the first one, we give a better definition of flexible public transport. The research gap is now clearly given in the next sentences. The objective of this study is to develop and apply a combinatorial method for strategic planning of flexible bus services. Recent studies have focused only on the operational level, aiming to develop services that can respond to demand in real time.

（2）The study employs a robust combinatorial approach, integrating quantitative social research and spatial analysis to gather both behavioral insights and spatial data relevant for route planning. While this approach is commendable, the methodology section could benefit from additional detail on the sample selection for the questionnaire survey, as well as on the spatial analysis techniques applied. Including more specifics on these aspects will enhance the transparency and reproducibility of the study's methodology.

Thank you for your insightful comment. To address the concern regarding sample selection, we have updated the last paragraph of section 2.1. The general strategy was to include as many different commuters with different preferences as possible. At the same time, we focused on the local community since these people are potential users of the service. This approach makes it possible to identify diverse needs, providing a good knowledge base for developing adaptive services based on those insights. In addition, in paragraph 8 of Chapter 4, where the limitations are presented, we underline that the application of the method requires the re-estimation of model beta parameters per rural area. Our study provides the framework to do so.

Regarding the spatial analysis techniques, it is an adjusted shortest path considering the land uses and the objective set from the conceptual framework. This was given in the last paragraph of section 2.2. Nevertheless, we decided to also highlight it in the first paragraph of this section 2.2.

（3）The analysis provides valuable insights into user preferences for flexible transport services, notably the preference for door-to-door service in the morning and stop-based service later in the day. This finding is insightful, yet further discussion on why this preference may exist (e.g., cultural or regional influences, work/school schedules) could deepen the reader's understanding. Additionally, while cost and time are highlighted as significant factors, more granularity on these influences (e.g., cost sensitivity across different income levels) could provide valuable context for policymakers.

Thank you for the good words. We tried to make some improvements in paragraph 2 of chapter 4: Discussion. So, the first finding is surely related to activities that are performed in each period of the day. This is underlined in the new version of the manuscript.

Regarding the second point, it is true that the price elasticity of this service is relatively high. This is valuable information for policymakers and potential investors, as there are substantial profit margins. Unfortunately, the income group was not proven as a statistically significant variable due to the colinearities with other socio-demographic characteristics. Therefore, the elasticity per income cannot be examined. However, in the updated Figure 6, we included gender and age group as extra dimensions of demand probability.

（4）The study introduces a novel approach to route planning by combining spatial analysis with multicriteria decision-making tools, an advancement that can benefit flexible transport planning. Nonetheless, the manuscript would be strengthened by offering a comparative analysis with other route-planning methodologies in similar contexts, highlighting the advantages and potential limitations of the proposed approach. This would position the study more clearly within the existing body of flexible transport literature and underscore its unique contributions.

This comment helped us to improve the discussion chapter. Thank you! In paragraph 1 of chapter 4, we highlight all the advantages of our approach: feasibility, applicability, and realistically flexible considering local needs and objectives. From the previous version of our paper, all the limitations were given in the 8th paragraph of chapter 4. After reviewing the paragraph, we are confident to say that all key points have been thoroughly highlighted.

Overall, the paper presents a well-structured and innovative study that contributes valuable insights to the field of flexible transportation planning. Addressing the comments above will enhance the clarity, depth, and applicability of the findings, ultimately making the study more impactful for both researchers and transportation planners.

Thank you very much for your thoughtful feedback. We hope these revisions will enhance the paper's contribution to the field and its value to both transport researchers and planners.

Reviewer 2 Report

Comments and Suggestions for Authors

This article mainly studies how to plan flexible bus service routes in regional cities and rural areas, while combining user preferences and spatial analysis. The manuscript has clear ideas, complete structure, and certain practical value. However, there are still some shortcomings in this article.

1. Has the questionnaire survey mentioned in the paper undergone predictive testing to ensure the effectiveness and reliability of the research? The authors should added the corresponding illustrations.

2. This article only analyzes the data through a binary logistic regression model. Suggest the author to add statistical indicators such as model fitting goodness for a more comprehensive analysis.

3. It is suggested that the author add a comparison of preferences for different user groups (such as age, gender, etc.) and an analysis of the potential value of flexible public transportation (such as cost savings, social benefits, etc.) in the results section.

Author Response

This article mainly studies how to plan flexible bus service routes in regional cities and rural areas, while combining user preferences and spatial analysis. The manuscript has clear ideas, complete structure, and certain practical value. However, there are still some shortcomings in this article.

Thank you for reviewing our article. We appreciate the positive feedback and simultaneously, we did our best to address all the shortcomings of the paper.

1. Has the questionnaire survey mentioned in the paper undergone predictive testing to ensure the effectiveness and reliability of the research? The authors should added the corresponding illustrations.

Thank you for your thoughtful question. Unfortunately, due to certain constraints, predictive testing is not possible in this research. Respondents gave stated preferences: "what they would do if". We did not have real preferences because the service does not exist today. It would not be methodologically correct to split the stated preference dataset (e.g., 70% - 30%) to test the model’s accuracy because these are not real-world data. Therefore, it is essential to gather post-introduction observations of the service to make valid comparisons and recalibrate the model. We mention this in the last paragraph of section 2.3, before providing probability prediction in Figure 6. Last, please keep in mind that the planning process is not strictly connected with the model. In route planning, we consult the model results to make critical decisions.

2. This article only analyzes the data through a binary logistic regression model. Suggest the author to add statistical indicators such as model fitting goodness for a more comprehensive analysis.

Thank you for this suggestion. We improved Table A.1 with the model in Appendix A to address this comment. We give the McFadden's ρ which was one statical indicator that was currently missing. At this point, we have to mention that there is no other way to analyze this dataset because we had designed the stated preferences experiment based on the utility function of the binary logit model.

3. It is suggested that the author add a comparison of preferences for different user groups (such as age, gender, etc.) and an analysis of the potential value of flexible public transportation (such as cost savings, social benefits, etc.) in the results section.

Thank you for this comment, since it leads to extra analysis and interesting results. We fully updated Figure 6 to show the probabilities for all gender and age groups. In addition, an estimation of total savings because of reduced waiting or walking time is now given in the second paragraph of section 3.1.

Thank you again for your thoughtful feedback.

Reviewer 3 Report

Comments and Suggestions for Authors

1-     The abstract did not show any novelty. It seems a project not a research paper.

2-     It is not clear in the abstract, what did the authors do in quantitative social research step.

3-     It is not clear in the abstract, how the results of this case study can be used for policymaking?

4-     Line 236, The reference has a problem.

5-     Please explain more about “Block Randomization” in line 249.

6-     Please explain about the relationship between the steps explained in lines 151-158 with stated preference survey.

7-     It is not clear, why you choose the equation in line 191. What was the reason? Why this type of modeling?

8-     Line 216, There are 324 cases by combining different input variables. How you collected such data? Explain more about fractional factorial design by an example. How fractional factorial design insures the sufficiency and coverage of data for the analysis? Namely, did you have all possible combinations in this method?

9-     Line 343, How you reached to these criteria?

1-  What is the relationship between these three steps? For example, how the outputs of step 1 can be used for the second step?

Comments on the Quality of English Language

1-     The abstract did not show any novelty. It seems a project not a research paper.

2-     It is not clear in the abstract, what did the authors do in quantitative social research step.

3-     It is not clear in the abstract, how the results of this case study can be used for policymaking?

4-     Line 236, The reference has a problem.

5-     Please explain more about “Block Randomization” in line 249.

6-     Please explain about the relationship between the steps explained in lines 151-158 with stated preference survey.

7-     It is not clear, why you choose the equation in line 191. What was the reason? Why this type of modeling?

8-     Line 216, There are 324 cases by combining different input variables. How you collected such data? Explain more about fractional factorial design by an example. How fractional factorial design insures the sufficiency and coverage of data for the analysis? Namely, did you have all possible combinations in this method?

9-     Line 343, How you reached to these criteria?

10-  What is the relationship between these three steps? For example, how the outputs of step 1 can be used for the second step?

 

Author Response

1-     The abstract did not show any novelty. It seems a project not a research paper.

2-     It is not clear in the abstract, what did the authors do in quantitative social research step.

“Thank you for your thoughtful feedback and for taking the time to review our work. Your comments have been invaluable in helping us enhance the quality of our paper. We have uploaded a revised version of the manuscript with all changes marked.

To address these comments jointly, we have improved the abstract to better highlight the novelty of our research and to clarify the steps undertaken in the quantitative social research phase.

The novelty of this study is related to the investigation of preferences for a future service. We considered the temporal dimension, which is very important in the process of developing adaptive (and customized) mobility solutions in rural areas. To the authors' knowledge, there is no other study that has followed this approach. The majority focus on planning on-demand services on a more operational level, preparing the right optimization algorithms to meet demand needs in real time. We focused on a strategiclevel. All the above are now given with few words in the first four sentences of the abstract.

3-     It is not clear in the abstract, how the results of this case study can be used for policymaking?

It is true! We cannot say that the results can help policymakers in other rural areas. But, the applied methodological approach can be proven helpful to planners and policymakers. We think that our study has some very interesting practical implications. For instance, Figure 12 can be a roadmap for creating customized solutions for public transport. To improve the abstract in this direction, we made some major changes in the last sentence.

4-     Line 236, The reference has a problem.

Thank you for the notice. This issue was fixed!

5-     Please explain more about “Block Randomization” in line 249.

Block randomization is actually an algorithm that distributes the four blocks of scenarios randomly. In other words, there was a single link for the survey, and the blocks were assigned randomly to the respondents. This ensures the same number of responses per block when the sample size increases. It also means the same number of observations per scenario, which leads to zero correlation among the independent variables in the final set. We provide more clarifications in the last paragraph of section 2.1.

6-     Please explain about the relationship between the steps explained in lines 151-158 with stated preference survey.

We believe there may have been a misunderstanding. The steps outlined in lines 174–180 refer to the process of designing the stated preference survey (or experiment). These steps were carefully followed to develop our survey. To clarify this point, we have updated the sentence in line 174 accordingly.

7-     It is not clear, why you choose the equation in line 191. What was the reason? Why this type of modeling?

Good question! Equation 1 is the mathematical transformation of Table 1. The selection of variables and variable levels is explained in detail in paragraphs 2 and 3 of section 2.1. There are multiple hypotheses and decisions before developing this function. Of course, all the hypotheses are tested by collecting and analyzing preferences. The literature that is presented in the Introduction contributed to defining the variables. Therefore, to address this comment, we made major modifications in these two paragraphs. We wanted to point out that Equation 1 comes as a result of Table 1.

8-     Line 216, There are 324 cases by combining different input variables. How you collected such data? Explain more about fractional factorial design by an example. How fractional factorial design insures the sufficiency and coverage of data for the analysis? Namely, did you have all possible combinations in this method?

First, we have to clarify that the fractional factorial design descreased the scenarios from 324 to 36. Therefore, 36 scenarios were integrated into the survey. We divided these 36 scenarios into 4 blocks so that each respondent would evaluate nine scenarios. This is, of course, more feasible.

The fractional factorial design ensures zero correlations among the variables (or multiplications of them) that are specifically included in Equation 1. What do we lose? We lose the chance the investigate the significance of extra interaction effects (multiplication of variables) after collecting the data. In other words, the survey is designed to explicitly estimate the model that is presented in Equation 1 and test all the mentioned hypotheses in paragraph 2 and 3 of section 2.1.

9-     Line 343, How you reached to these criteria?

We consulted the literature and basically previous studies that designed a flexible service in the strategic level. Also, the criteria are related to the strategic objectives of this service. These were set in the beginning of the project. In the fourth paragraph of section 2.3, we give better explanations about how the MCA criteria were selected.

Thank you again for your thoughtful feedback.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Thanks for the efforts of responding my comments. You have addressed my concerns to a satisfactory level.