AI-Driven Approach for Enhancing Sustainability in Urban Public Transportation
![](/bundles/mdpisciprofileslink/img/unknown-user.png)
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis is a valuable study. However, there are some questions that need to be clarified.
The parameters between lines 401 and 429 are important for the research. How are these parameters taken? And how to ensure that these values are valid?
What are the data in Table 4? Is there a definite physical meaning? Where is the improvement of optimization reflected?
How dose the analysis of lines 447 to 474 arrive at? In my opinion, the largest values in the three cities in Table 4 are passenger satisfaction.
Author Response
Answers to second reviewer:
- The parameters between lines 401 and 429 are important for the research. How are these parameters taken? And how to ensure that these values are valid?
Answer: all of these parameters have been used to conduct q value algorithm for reinforcement learning, all steps displayed have been used to ensure model validity, and also model robustness has been checked according to a request by another reviewer.
2. What are the data in Table 4? Is there a definite physical meaning? Where is the improvement of optimization reflected?
Answer: Data displayed is reflecting on results of conducting Qvalue algorithm of reinforcement learning. Out of max value 100, research results for all dimensions in three sampled cities are the ones displayed in table 4, in the form of total rewards parameter. Improvements are represented through values of total reward parameter for each dimension and for each city. As it was already mentioned in the text, the total reward parameter suggest that the system is learning to balance the key dimensions well, as the rewards are consistently above a baseline, indicating successful optimization of the public transportation parameters.
3. How dose the analysis of lines 447 to 474 arrive at? In my opinion, the largest values in the three cities in Table 4 are passenger satisfaction.
Answer: Results have been commented in detail after table 4, and now the analysis comments have been broadened (also on demand made from another reviewer).
Reviewer 2 Report
Comments and Suggestions for AuthorsDear Authors,
"I commend you on your valuable work and the significant effort invested in enhancing public transport across three cities in Serbia using AI technologies. However, I noticed some unacceptable inconsistencies in the references—some were missing, and others appeared irrelevant to the paper's topic. It is therefore essential to correct the reference section and address the remaining comments provided."
Kindest regards,
Comments for author File: Comments.pdf
Author Response
Please find all answers to reviewers comments (in the revised version it can be followed through track changes:
- The sentence in abstract regarding future research has been deleted from abstract. Academic editor required another sentence to be defined in end of abstract.
- the authors gave their detailed view of the term "sustainability" right after defining research question
- reference no 9 has been corrected
- reference no 10 has been corrected
- reference no 11 has been corrected
- replacement for word "health" has been provided.
- references no 12 and 13 have been substituted.
- Comment for reference 17 has been checked and corrected now.
- Comment for reference 18 has been checked and corrected now.
- reference no 12 has been corrected
- All mentioned references (21-57) have been corrected and replaced, the error happened through the tool for automatic referencing so it has now been corrected.
- reference no 58 has been replaced.
- All mentioned references (59-66) have been corrected and replaced, the error happened through the tool for automatic referencing so it has now been corrected.
- existing statistical measures for reducing headway regulations were not the main subject of this paper, since AI application was the main focus.
- mentioned reference is replaced with proper one (from mdpi). sustainability metrics are hard to define and are not the topic of this paper.
- table 1 has been updated with proper format
- all references in table 2 have been corrected and replaced.
- the authors are grateful for the proposal and will work on second and third paper, with more detail for proposed paper topics. In this version they see it as an integral topic that cannot be shortened or divided.
- this sentence in framework for analysis has been changed.
- mentioned function reward is a precondition for conducting Q algorithm
- table 3 name has been defined. Yes it forms all research dimensions in three 3 steps (S).
- table 4 name has been defined, top ceiling for rewards is 100 and it has now been mentioned
- authors added specific part of the paper related to comparing results between 3 cities. the paragraph starts with: " Belgrade demonstrated high operational efficiency " and "Overall, the impact of AI-driven optimizations varied across the three cities..."
- model robustness has been checked and appropriate text has been added.
- this limitation can be overcome with synthetic data generation and this has been mentioned now in the text covering limitations
- repeated statements have been corrected
The authors are willing to implement any further changes
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsAll my comments are responsed. I do not have any other questions.