An Open-Source System for Public Transport Route Data Curation Using OpenTripPlanner in Australia

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript presents an open-source implementation of a system for planning routes that combine public transportation and walking, aiming to optimize both travel time and distance. The current implementation is limited to major cities in Australia.

Unlike commercial solutions, the proposed system supports programmatic interaction via an API and enables bulk processing. It provides three interfaces: (1) an interactive web-based map; (2) a programmatic HTTP API; and (3) a bulk-processing interface.

The system was evaluated by comparing its results with a benchmark commercial solution in two dimensions: route correctness and area coverage. Additionally, the authors assessed server performance and conducted a local benchmark.

The authors also provide a web page to test the system. This webpage generally performs well, although at times it fails to display the map or routing options. Nevertheless, its availability is acknowledged by this reviewer.

However, I have several questions that I would like the authors to address:

- The system is currently restricted to major cities in Australia. What would be required to generalize it? For instance, could the system be deployed in Europe or the United States, and if so, how? Addressing this issue in Section 4 would substantially strengthen the paper.

- The Discussion section could be improved. For example, Figures 11 and 12—and the paragraph in which they are discussed—are not clear. The authors should clarify this section and explain the relevance of these figures.

- The data presented in Table 7 would be much clearer if shown as a plotted chart, which would provide a more intuitive visualization of how the number of clients impacts system performance.

Minor issues:

- Section 2 should be titled “Related Work”, not “Related Works.”

- Are references 8 and 19 essentially the same? If so, the authors should revise the reference list to avoid duplication.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The authors provide a reasonable review of the existing literature, noting that many prior studies focus on static attributes and overlook dynamic characteristics of transit networks. However, the literature base remains relatively limited, and some items are not fully elaborated. Additionally, it is unclear what differentiates position 8 from position 19 in the manuscript.

The paper introduces OpenTripPlanner (OTP) as an open-source routing engine and implements an OTP 1.5-based multi-router system tailored to Australia’s federated transport data landscape. The system, publicly accessible at https://ptplanner.latrobe.edu.au, supports route planning across eight metropolitan regions. This is a positive contribution from an engineering perspective, and the development of an adaptable, open-source infrastructure for large-scale public transport analysis is promising.

However, as a scientific article, the manuscript is insufficient. It contains extensive implementation details, which resemble a user manual rather than a research paper. The comparison of OTP outputs with Google Maps lacks methodological rigor, and the statement regarding city-wide routing tests is unclear. The experiment description is vague (e.g., “containing numerous trip plans in several cities at different times of the day”), and Figure 9 is poorly readable (it should be improved to ensure clarity and interpretability of the presented data).

The server performance evaluation is more interesting scientifically, as it addresses scalability under stress conditions. Yet, the analysis is incomplete: MedianLatency is missing from Table 7, and variability measures such as standard deviation should be included. 

In summary, while the study demonstrates the potential of OTP as a platform for large-scale public transport analysis, the current manuscript adds little to the advancement of the field and remains more of an engineering curiosity. Perhaps in the future, if an improved algorithm is developed and implemented, the work could be published as a more substantial contribution. At present, the paper does not meet the criteria for a scientific publication due to the lack of methodological rigor, insufficient analytical depth, and limited theoretical contribution.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The paper is generally clear, well structured, and offers meaningful contributions through system design, implementation, and performance evaluation. This work has practical value and demonstrates solid reproducibility. However, several aspects could be improved to enhance the scientific depth, clarity, and contextual positioning of the study. I outline major and minor comments below.

1. The comparison against Google Maps is purely qualitative. It would strengthen the analysis to quantify differences in travel time, route length, or transfer count, analyze systematic biases in OTP results (e.g., tendency to select bus-only routes), and discuss the implications of using static GTFS schedules.

2. While the paper reports that OTP 1.5 requires substantial RAM in multi-router mode, the authors should analyze memory consumption per city graph and explain how performance would scale with additional cities or larger GTFS datasets.

3. The related work section would benefit from deeper engagement with recent accessibility studies using OTP or GTFS-based routing engines.

4. The paper occasionally uses “route” and “router” interchangeably, please consider standardizing terms.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments and Suggestions for Authors

This manuscript addresses an important and timely topic in public transport accessibility and open-source routing systems. While the paper demonstrates strong technical implementation and reproducibility, some sections require substantial refinement to meet the expectations of our journal. The comments below are organized by section and aim to provide constructive, detailed feedback that highlights both strengths and areas for improvement. They emphasize the need to move beyond technical reporting toward a more analytical, theory-driven, and policy-relevant framing that situates the work within global discourses on transport justice, accessibility, and sustainable urban planning.

Abstract

The abstract spends too much space on implementation details (multi-threaded execution, router separation, benchmarking). It is also expected that abstracts emphasize novelty, impact, and broader implications. Accessibility and equity are mentioned, but the abstract does not clearly articulate how this system advances transport justice or sustainable urban planning. Phrases like “trade-offs between local and cloud setups” or “practical limitations” are mentioned without concrete findings (e.g., performance ceiling, coverage rates). Keywords could be expanded: Missing terms like Transport Equity, Urban Planning, Accessibility Analysis, which would increase visibility in indexing.

Emphasize how this system contributes to equitable transport planning and policy-relevant research, not just technical deployment. Add quantitative highlights; mention key results (e.g., “achieved 91–99% coverage in most cities, with performance up to 3,000 requests/minute”). Reduce technical jargon; move details like multi-threaded execution or router separation to the methodology section. Strengthen keywords; Add terms related to equity, sustainability, and urban mobility.

Introduction

The introduction reads more like a technical background than a critical framing of the research gap. A sharper articulation of why existing methods are insufficient and what theoretical or methodological innovation this paper brings is expected. While equity and sustainability are mentioned, the introduction does not connect to global discourses in transport justice, accessibility theory, or sustainable urban planning. The listed contributions emphasize system deployment and performance evaluation, but lack emphasis on scientific novelty (e.g., advancing accessibility metrics, reproducibility frameworks, or methodological transparency). The introduction cites GTFS and OTP but does not sufficiently reference foundational works in accessibility analysis (e.g., Geurs & van Wee, Litman, Curtis & Scheurer). The text jumps quickly from GTFS challenges to OTP deployment without fully developing the research gap.

Explicitly state what is missing in current PT accessibility studies (e.g., lack of reproducible routing frameworks, limited national-scale analysis, absence of equity-focused metrics). Link the study to broader frameworks such as transport justice, spatial equity, and sustainable accessibility. Reframe contributions; instead of listing technical achievements, highlight methodological and societal contributions (e.g., “This work advances reproducible accessibility analysis by demonstrating an open-source, multi-router framework applicable to federated transport systems”). Strengthen literature integration by adding references to seminal works in accessibility and transport equity to show awareness of the global research landscape. Improve narrative flow by moving technical details (multi-threading, router separation) to methodology, and keep the introduction focused on framing the problem and research significance.

Related Works

The section reads as a list of studies rather than a synthesized review. A critical synthesis that identifies gaps and positions the current work clearly is expected. While equity and accessibility are mentioned, the section does not engage with foundational theories (e.g., Geurs & van Wee’s accessibility framework, Litman’s transport equity, Curtis & Scheurer’s spatial network analysis). The discussion of scraping is detailed but lacks critical reflection on ethical, legal, and methodological implications. The section ends abruptly with OTP deployment, without clearly articulating how the gaps in the literature justify the chosen approach.

Synthesize, don’t list; instead of enumerating studies, group them into themes (e.g., static GTFS analyses, equity-focused studies, scraping approaches, open-source routing tools). Highlight the research gap; explicitly state that while GTFS and scraping have been widely used, there is a lack of reproducible, open-source, national-scale routing frameworks, especially in federated data contexts like Australia. Engage with theory by adding references to seminal works in accessibility and equity to strengthen the conceptual foundation. Discuss ethics and reproducibility; reflect on the limitations of scraping (licensing, transparency, reproducibility) and how OTP addresses these issues.

Methodology or OTP Multi-Router Architecture

The section reads like a manual or deployment guide. It is expected that the methodology to balance technical reproducibility with conceptual framing. The description focuses on implementation rather than methodological contribution. It does not explain how this approach advances accessibility analysis compared to existing methods. The methodology is presented in isolation, without linking back to the research questions (e.g., how multi-router architecture enables equity-focused analysis). Heavy reliance on OTP 1.5 (deprecated multi-router support) seems to be a vulnerability. The paper should discuss the implications of using outdated software and potential migration paths. Figures remain underutilized; they could be used to highlight methodological advantages or limitations.

Reframe methodology as research design instead of presenting OTP deployment as a technical manual, and emphasize how the chosen architecture supports the study’s goals (e.g., reproducibility, scalability, equity analysis). Clarify what is new here (e.g., adapting OTP multi-router for federated GTFS datasets at national scale, enabling reproducible accessibility studies. Explicitly connect each methodological choice (multi-router, bulk processing, API) to research questions (e.g., coverage analysis, equity evaluation). Discuss limitations by acknowledging that OTP 1.5 is outdated and explain why it was chosen, while outlining how future work could adapt to OTP 2.x. Improve figures by using diagrams not only to show architecture but also to compare efficiency, scalability, or coverage between single and multi-router setups.

Data Preparation & Implementation

Much of the section reads like a deployment manual. Methodology should emphasize research design and scientific contribution, not just technical instructions. The section does not explain why certain choices (e.g., trimming OSM data, setting walking speed to 1.33 m/s, limiting walk distance to 1000 m) are methodologically significant for accessibility analysis. GTFS feeds are assumed reliable, but issues like missing routes, inconsistent calendars, or agency-level differences are not critically discussed. The section does not explicitly connect data preparation choices to the study’s goals (e.g., equity analysis, reproducibility, scalability). Figures (map trimming, folder structure, configuration files) show process steps but do not highlight methodological insights or comparative advantages.

Reframe technical steps as methodological choices. For example, explain how trimming OSM data improves computational efficiency and accuracy in accessibility studies. Discuss data quality and limitations; critically evaluate GTFS feeds (e.g., coverage gaps, multimodal inconsistencies) and how these affect results. Justify parameter settings by providing a rationale for walking speed, maximum walking distance, and mode restrictions, linking them to established accessibility literature. Connect to objectives; explicitly state how the chosen data preparation pipeline supports reproducibility, scalability, and equity-focused analysis. Enhance figures instead of only showing folder structures; include analytical diagrams (e.g., data flow, error rates, or comparative coverage maps).

Evaluation Section

Results are reported but not deeply interpreted. The evaluation must explain why discrepancies occur and what they mean for accessibility and equity. Google Maps comparison seems to be superficial; the analysis notes differences in travel times and modes but does not critically assess the implications (e.g., how reliance on static GTFS limits multimodal realism, or how equity is affected by missing transfers). Coverage analysis lacks a socio-spatial dimension; success/failure rates are reported, but there is no discussion of which communities are underserved (e.g., low-income suburbs, peripheral areas). Server performance evaluation is too technical; while detailed, it does not connect to research relevance (e.g., how scalability affects reproducibility of national accessibility studies). Figures are also underutilized; route visualizations and correlation matrices are presented but not analyzed in depth.

Deepen the interpretation of results. For route correctness, explain how OTP’s reliance on static GTFS leads to equity-relevant limitations (e.g., bus-only routes in cities where multimodal options exist). Link coverage failures to geographic and demographic characteristics (outer suburbs, low-density areas, marginalized communities). This would elevate the paper’s contribution to transport justice. Discuss how server scalability affects institutional adoption (e.g., universities, government agencies) and the reproducibility of large-scale studies. Strengthen comparative framing instead of only contrasting OTP vs Google Maps, highlight the trade-off between transparency/reproducibility (OTP) and real-time adaptability (Google). Use figures analytically. For example, interpret coverage maps to show inequities, or use correlation matrices to explain bottlenecks in server performance.

Discussion Section

The section mostly repeats results instead of interpreting their broader implications for accessibility, equity, or sustainable transport planning. No engagement with transport justice, accessibility frameworks, or sustainability debates. This makes the discussion feel more like a technical report than a high-level research paper. Coverage failures in Brisbane and Darwin are noted but not analyzed in terms of socio-spatial inequities (e.g., underserved communities, suburban disadvantage). The discussion does not connect findings to planning or policy implications (e.g., how open-source routing could support government agencies, universities, or NGOs). Future work is narrow; suggestions are limited to technical upgrades (OTP 2.x, bulk-processing). Broader research directions (integration with socio-economic data, equity metrics, climate adaptation) are absent.

Deepen analytical framing by interpreting discrepancies between OTP and Google Maps as evidence of the trade-off between reproducibility and real-time adaptability, linking this to equity and accessibility debates. Integrate theory by explicitly connecting findings to frameworks such as Geurs & van Wee’s accessibility dimensions (availability, accessibility, affordability, acceptability) or Litman’s transport equity. Highlight socio-spatial inequities by discussing how coverage failures disproportionately affect peripheral or low-density communities, and what this means for transport justice. Show how open-source routing systems can empower public institutions, reduce reliance on proprietary APIs, and support equitable planning. Expand future work by suggesting integrating socio-economic datasets, developing equity-focused accessibility metrics, and exploring applications in climate-resilient transport planning.

 

Conclusion

The conclusion is too technical and narrow. It focuses on performance metrics (3,000 RPM, resource intensity) rather than broader societal and policy implications. It also does not connect findings to transport justice, accessibility theory, or sustainable urban planning. Future work is still underdeveloped because suggestions are mostly technical (OTP 2.x, diagnostics). Missing broader research directions (integration with socio-economic datasets, equity-focused metrics, climate adaptation).

Reframe the conclusion for impact. Emphasize how the system contributes to equitable transport planning, reproducibility, and sustainable urban mobility. Link findings to accessibility frameworks and transport justice debates. Suggest integration with socio-economic data, development of equity metrics, and application to climate-resilient transport planning.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

In my opinion, the manuscript has been improved enough to warrant publication in Computers. However, despite the improvements made, there remain some issues that should be addressed.

The authors revised the manuscript to address concerns about limited theoretical and methodological depth. They clarified the purpose of each evaluation—route correctness, coverage, and server scalability—and added discussion on discrepancies between static GTFS data and real-time multimodal platforms. The theoretical framing was strengthened by explaining how the platform can support future equity-focused and sustainable transport studies. They incorporated references to established frameworks, such as those by Geurs & van Wee and Litman, and outlined potential integration with socio-demographic data in the discussion and future work sections. 
Methodological clarity was improved by reducing low-level implementation details and presenting configuration choices as literature-based methodological decisions. The authors emphasized that the platform is reproducible, scalable, and openly accessible, providing a practical foundation for transport accessibility research. 
They acknowledged the engineering focus but positioned the work as a stepping stone toward more advanced algorithmic and theoretical contributions. Future plans include integrating real-time data, enabling intermodal routing, and applying the system to policy-relevant research, such as access to essential services. 
Overall, the revision aims to balance practical implementation with theoretical relevance and demonstrate the platform’s potential for inclusive and sustainable transport studies.

The remarks below concern minor issues that could be addressed for improved clarity:
1. The data presented in Table 2 should be commented on in greater detail, and it would be helpful to clarify the meaning of the symbols |E| and |V|—presumably the number of edges and vertices in the graph, but this should be stated explicitly 
2. While Table 7 reports successful and failed route responses per city, the meaning of ‘failure’ in this context requires clarification.
3. The authors state: ‘Standard deviation values were not included, as the differences between mean and median latency across all test cases were consistently small (within ±0.05 seconds), indicating low variability and a stable response pattern.’
I disagree with the assumption that a small difference between the mean and median indicates low variability and a stable response pattern. For example, in a sample such as {0, 5, 10}, the mean and median are identical, yet the data are highly dispersed. A small difference between these measures only suggests that the distribution is approximately symmetric, not that variability is low. Therefore, omitting standard deviation values limits the reader’s ability to assess variability accurately. However, the authors also note that ‘no significant outliers observed’, which may not be a very precise statement, but it could be sufficient in this context, and explains why the authors did not include Standard deviation values.
4. The value of 0.4 reported as a failure in the first row of Table 9 is somewhat puzzling and would benefit from clarification.
5. Computing a correlation between successes and failures is conceptually problematic, as these quantities are complements; intuitively, more successes necessarily imply fewer failures, yielding a mechanically strong negative association that is not informative. It is therefore surprising that Figure 9 reports a correlation of 0.16, which warrants clarification.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

This revised manuscript demonstrates clear progress compared to the earlier version. The authors have incorporated several important improvements, including:

• Addition of quantitative highlights in the abstract (coverage rates, throughput).
• Expanded keywords to include equity and urban planning.
• Integration of seminal references (Litman, Geurs & van Wee, Curtis & Scheurer).
• More explicit discussion of coverage gaps in Brisbane and Darwin, linked to underserved areas.
• Future work now includes socio-economic integration and equity-focused accessibility metrics.

These changes strengthen the paper considerably. However, some areas still require deeper refinement to fully meet the expectations of a Q1 journal:

Abstract

The inclusion of quantitative results and equity-related keywords is commendable. However, the abstract still emphasizes technical feasibility more than theoretical or policy relevance. Please clarify how the platform advances transport justice and sustainable urban mobility, beyond performance metrics.

Introduction

The revised introduction now mentions social inclusion, sustainability, and climate resilience. Yet, the framing remains largely technical. Stronger integration of the newly cited foundational works (Litman, Geurs & van Wee, Curtis & Scheurer) would help position the study within global discourses on accessibility and equity. Contributions should highlight methodological and societal novelty, not only technical deployment.

Related Work

The section is improved, with clearer thematic grouping and discussion of scraping ethics. Still, the theoretical synthesis is limited. The cited works on accessibility and equity should be explicitly connected to the research gap, showing how this study advances beyond prior approaches.

Methodology & Data Preparation

The authors now acknowledge limitations (e.g., OTP 1.5, fragmented GTFS feeds, cross-border routing issues). This is a valuable addition. However, the section still reads like a technical manual. Please reframe methodological choices as research design decisions, explicitly linking them to reproducibility, scalability, and equity analysis. Parameter choices (walking speed, maximum distance) should be justified with reference to accessibility literature.

Evaluation

The coverage analysis now highlights socio-spatial gaps in Brisbane and Darwin. This is a strong step toward equity framing. However, interpretation remains limited. Please connect these findings more explicitly to transport justice debates, e.g., how underserved communities are disadvantaged by current data structures. The comparison with Google Maps is still framed mainly in technical terms. A deeper discussion of the trade-off between reproducibility (OTP) and adaptability (Google Maps) would elevate the contribution.

Discussion

The revised discussion acknowledges underserved regions and proposes equity-focused future work. This is a major improvement. Still, the section does not fully leverage the newly added references. Please explicitly apply frameworks such as Geurs & van Wee’s four dimensions of accessibility or Litman’s distributional equity to interpret your results. Policy implications should be expanded: how can open-source routing empower agencies, universities, or NGOs to pursue equitable planning?

Conclusion

The conclusion now mentions equity-focused research directions, which is positive. However, it still emphasizes technical feasibility. Please reframe to highlight broader societal impact: reproducibility, transparency, and contribution to sustainable urban mobility. Explicitly connect findings to transport justice and accessibility theory to leave a stronger impression.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 3

Reviewer 4 Report

Comments and Suggestions for Authors

The revised manuscript shows clear and meaningful improvement compared to the previous version. I appreciate the authors’ efforts in strengthening the conceptual framing, expanding the discussion of equity and accessibility, and integrating foundational references such as Geurs & van Wee. The addition of the four-component accessibility framework in the Discussion section is particularly valuable and significantly enhances the theoretical grounding of the paper. The expanded interpretation of coverage gaps and the clearer articulation of the platform’s reproducibility benefits also contribute positively to the manuscript. That said, some areas still require further refinement to fully meet the expectations of a high-impact journal:

1. Strengthen theoretical integration across the manuscript. While the Discussion now engages well with Geurs & van Wee’s framework, the Introduction and Related Work sections still rely primarily on descriptive or technical framing. More explicit integration of accessibility and transport justice theories throughout the narrative would help position the contribution more clearly within global scholarly debates.
2. Deepen the equity-oriented interpretation of results. The coverage analysis now acknowledges underserved areas, but the interpretation remains largely descriptive. The manuscript would benefit from a more explicit socio-spatial analysis, linking observed gaps to broader issues of transport disadvantage, peripheral urbanisation, or demographic vulnerability.
3. Clarify the policy relevance of the performance evaluation. The system performance section remains highly technical. The authors should more clearly articulate why scalability, latency, and throughput matter for institutional adoption, reproducible research, and public-sector planning workflows. Connecting these findings to practical use cases would strengthen the contribution.
4. Reduce technical detail in the methodology and shift toward research design. Some subsections (e.g., folder structures, API examples, configuration details) still read like deployment documentation. These could be streamlined or moved to supplementary materials, allowing the main text to focus on methodological rationale and analytical implications.
5. Enhance the analytical use of figures. Several figures (e.g., route visualisations, coverage maps) are currently descriptive. Adding brief analytical commentary highlighting spatial patterns, inequities, or methodological insights would improve their contribution to the argument.
6. Expand the Conclusion to emphasise societal and policy impact. The revised conclusion is stronger, but it still leans toward technical feasibility. Highlighting how the platform can support equitable planning, transparent decision-making, and sustainable mobility strategies would leave a stronger final impression.

Overall, the manuscript is progressing well and has strong potential. With deeper theoretical integration, clearer policy relevance, and more analytical interpretation of results, it can make a valuable contribution to research on public transport accessibility and open-source routing systems.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf