Next Article in Journal
Enhancing Employee Well-Being Through Scene Innovation in Retail Enterprises: A Case Study on the Chinese Enterprise Pang Donglai
Previous Article in Journal
Sustainable Integrated Algal Biomass Biorefinery: Synergistic Macronutrient Optimization and Electro-Flocculation Coagulation Harvesting
 
 
Article
Peer-Review Record

Optimizing the Operational Scheduling of Automaker’s Self-Owned Ro-Ro Fleet

Sustainability 2025, 17(19), 8683; https://doi.org/10.3390/su17198683
by Feihu Diao, Yijie Ren and Shanhua Wu *
Reviewer 1: Anonymous
Reviewer 2:
Reviewer 3: Anonymous
Reviewer 4: Anonymous
Sustainability 2025, 17(19), 8683; https://doi.org/10.3390/su17198683
Submission received: 8 August 2025 / Revised: 19 September 2025 / Accepted: 24 September 2025 / Published: 26 September 2025
(This article belongs to the Section Sustainable Transportation)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper addresses a highly relevant and timely topic concerning the operational scheduling of self-owned Ro-Ro (Roll-on/Roll-off) fleets for automakers, with a specific focus on the burgeoning Chinese New Energy Vehicle (NEV) export market. The authors effectively highlight a critical research gap: the existing literature predominantly focuses on the profit-maximizing objectives of traditional shipping companies, overlooking the unique operational needs of cargo-owners (automakers) who prioritize meeting export demands while minimizing costs. To address this, the study proposes a mixed-integer programming model that optimizes fleet operational costs under deterministic demand, considering decisions such as port selection, sailing speed, and loading quantities. A genetic algorithm is designed to solve the model, and its effectiveness is demonstrated through a case study. The work makes a valuable contribution by providing a decision-making tool for automakers to manage their own maritime logistics.

1.  The model formulation section requires careful proofreading. There appears to be an inconsistency between the variable definitions in the tables and their usage in the equations. For instance, Table 3 defines a binary decision variable as x_{ky}, but the objective function and constraints use notations like x_{kj} and x_{kij}. Please ensure that all variables, parameters, and sets are consistently and clearly defined and used throughout the entire paper. This is a fundamental requirement for a quantitative study.

2.  The paper's introduction accurately points out that automakers' shipping demands are "heavily shaped by market uncertainty". However, the model formulation is based on a "premise of deterministic automobile export transportation demands". This is a significant simplification that contradicts the problem description. The authors should either justify this assumption more thoroughly or, ideally, extend the model to incorporate demand uncertainty using a stochastic programming or robust optimization approach. This would substantially enhance the practical applicability and academic contribution of the work.

3.  While the literature review is comprehensive, it could benefit from a more in-depth critical analysis. The authors successfully identify the research gap related to cargo-owner-focused models. However, a stronger discussion on why existing models (e.g., those for bulk carriers) are unsuitable for the complexities of Ro-Ro vehicle transport would strengthen the argument for the proposed model. A more detailed comparison with relevant studies could further highlight the novelty of the present work. To provide some additional background on the application of heuristic algorithms, it is recommended to refer to the introduction to the application in the article Bioinspired Discrete Two-Stage Surrogate-Assisted Algorithm for Large-Scale Traveling Salesman Problem.

4.  The conclusion is concise but could be expanded to provide a more impactful summary of the findings and their implications. The authors should explicitly discuss the limitations of the current study, such as the deterministic assumption, and clearly outline potential future research avenues. For example, future work could explore incorporating fleet expansion decisions, considering multi-commodity transport (e.g., with non-vehicle cargo), or integrating the model with real-time data for dynamic scheduling.

5.  The abstract is quite long and could be more concise. It should be streamlined to immediately grab the reader's attention by stating the specific problem, the proposed method, and the key contributions in a more direct manner. Cutting down on background information and focusing on the core findings would improve its readability and effectiveness.

6.  There are several minor formatting and grammatical issues. For example, the objective function in Equation 1 seems to have some missing summation indices and possibly other notational errors. I recommend a thorough check of all mathematical expressions and a full proofread of the manuscript by a native English speaker to catch any linguistic or stylistic errors.

Author Response

Comment 1: The model formulation section requires careful proofreading. There appears to be an inconsistency between the variable definitions in the tables and their usage in the equations. For instance, Table 3 defines a binary decision variable as x_{ky}, but the objective function and constraints use notations like x_{kj} and x_{kij}. Please ensure that all variables, parameters, and sets are consistently and clearly defined and used throughout the entire paper. This is a fundamental requirement for a quantitative study.

Response 1: Thank you for pointing out the variable notation inconsistency. We have carefully proofread the Model Formulation section, unified all variable notations, and verified consistency between variable definitions in tables and their usage in objective functions/constraints. All sets, variables, and parameters now maintain clear and uniform expressions throughout the paper.

Comment 2: The paper's introduction accurately points out that automakers' shipping demands are "heavily shaped by market uncertainty". However, the model formulation is based on a "premise of deterministic automobile export transportation demands". This is a significant simplification that contradicts the problem description. The authors should either justify this assumption more thoroughly or, ideally, extend the model to incorporate demand uncertainty using a stochastic programming or robust optimization approach. This would substantially enhance the practical applicability and academic contribution of the work.

Response 2: Thank you for your insightful comment on demand uncertainty. We acknowledge this as a key limitation of the current study and have addressed it in two ways:

First, we have highlighted extending the model to stochastic/robust optimization (to incorporate demand volatility) as a core future research direction in the Conclusion.

Second, while the model is based on deterministic demand, its flexibility to adapt to demand changes is verified in the Case Study’s sensitivity analysis—adjusting demand parameters (e.g., European market demand fluctuations) yields corresponding optimized scheduling schemes, demonstrating practical adaptability.

We appreciate this feedback, which will guide our follow-up work to enhance the model’s realism.

Comment 3: While the literature review is comprehensive, it could benefit from a more in-depth critical analysis. The authors successfully identify the research gap related to cargo-owner-focused models. However, a stronger discussion on why existing models (e.g., those for bulk carriers) are unsuitable for the complexities of Ro-Ro vehicle transport would strengthen the argument for the proposed model. A more detailed comparison with relevant studies could further highlight the novelty of the present work. To provide some additional background on the application of heuristic algorithms, it is recommended to refer to the introduction to the application in the article Bioinspired Discrete Two-Stage Surrogate-Assisted Algorithm for Large-Scale Traveling Salesman Problem.

Response 3: Thank you for your valuable comments on the literature review. We have addressed your suggestions as follows:

(1) Regarding the suitability of bulk carrier models for Ro-Ro transport: We acknowledge our oversight in not explicitly clarifying this. Ro-Ro vehicle transport (a form of liner shipping) differs fundamentally from bulk carrier transport in cargo characteristics (finished vehicles vs. bulk goods), loading/unloading methods (ramp-driven vs. bulk handling), and vessel operation modes—these are common maritime logistics knowledge, which led us to omit detailed discussion. We have supplemented a brief note on this distinction in the literature review to strengthen the rationale for our proposed model.

(2) The recommended articles have been cited in the literature review section to enrich the background on heuristic algorithm applications.

Comment 4: The conclusion is concise but could be expanded to provide a more impactful summary of the findings and their implications. The authors should explicitly discuss the limitations of the current study, such as the deterministic assumption, and clearly outline potential future research avenues. For example, future work could explore incorporating fleet expansion decisions, considering multi-commodity transport (e.g., with non-vehicle cargo), or integrating the model with real-time data for dynamic scheduling.

Response 4: Thank you for your comment on the conclusion. We have expanded the Conclusion section as suggested: it now includes a more impactful summary of key findings, detailed theoretical and practical implications, explicit discussion of limitations, and clear future research avenues. These revisions enhance the conclusion’s depth and guidance.

Comment 5: The abstract is quite long and could be more concise. It should be streamlined to immediately grab the reader's attention by stating the specific problem, the proposed method, and the key contributions in a more direct manner. Cutting down on background information and focusing on the core findings would improve its readability and effectiveness.

Response 5: Thank you for your comment on the abstract. We have streamlined it as suggested: cut redundant background information, directly highlighted the specific problem, proposed method, core findings, and key contributions. The revised abstract is more concise and focused, improving readability and effectiveness.

Comment 6: There are several minor formatting and grammatical issues. For example, the objective function in Equation 1 seems to have some missing summation indices and possibly other notational errors. I recommend a thorough check of all mathematical expressions and a full proofread of the manuscript by a native English speaker to catch any linguistic or stylistic errors.

Response 6: Thank you for your comment on formatting and grammar. We have addressed the issues as follows:

(1) Thoroughly checked all mathematical expressions;

(2) Had the entire manuscript proofread by a native English speaker to fix linguistic and stylistic errors.

Reviewer 2 Report

Comments and Suggestions for Authors

This paper presents the optimal scheduling of the RO-RO fleets to minimize the total cost while satisfying the demands with port-related constraints. The authors must provide the originality of their mathematical models in comparison with the existing literature (e.g., airplane fleet management). Their mathematical model "seems" to be almost identical with the existing model in mathematical perspectives. Of course, this paper may be original in the aspect of applications for RO-RO fleet management. However, in order to claim merits for the application research, stochastic nature of the RO-RO fleet operations must be addressed while the current model is completely deterministic without consideration of stochastic demands and ETAs, etc.

Author Response

Comment 1: This paper presents the optimal scheduling of the RO-RO fleets to minimize the total cost while satisfying the demands with port-related constraints. The authors must provide the originality of their mathematical models in comparison with the existing literature (e.g., airplane fleet management). Their mathematical model "seems" to be almost identical with the existing model in mathematical perspectives. Of course, this paper may be original in the aspect of applications for RO-RO fleet management. However, in order to claim merits for the application research, stochastic nature of the RO-RO fleet operations must be addressed while the current model is completely deterministic without consideration of stochastic demands and ETAs, etc.

Response 1: Thank you for your insightful comment. We acknowledge that the originality of this study lies in practical application to Ro-Ro fleet management rather than mathematical model innovation. While the model adopts a deterministic framework, it is meaningful for automakers’ real-world needs: it targets the understudied scenario of automaker-owned Ro-Ro fleets and generates actionable scheduling schemes (validated via case study) to solve cost-control and demand-fulfillment challenges in NEV export logistics.

The theoretical and practical implications of this study lie in:

First, it enriches the theoretical system of Ro-Ro shipping scheduling by constructing a targeted optimization model for automaker’s self-owned fleets, effectively filling the research gap in scheduling optimization for non-third-party Ro-Ro fleets and providing a new theoretical framework for subsequent studies on self-owned maritime logistics assets;

Second, the integration of multi-dimensional decision variables (port sequence, load, speed) into the mixed-integer programming model expands the application scope of such models in maritime logistics, offering a valuable reference for optimizing complex scheduling problems with multiple constraints.

Third, for NEV manufacturers, particularly Chinese manufacturers engaged in overseas expansion, the proposed scheduling optimization method can directly support the reduction of operational cost for self-owned Ro-Ro fleets. Specifically, by optimizing port call routes and sailing speeds, enterprises can enhance fleet utilization and alleviate cost pressures arising from high shipping rates of third-party Ro-Ro services; for logistics practitioners in the Ro-Ro shipping industry, the methodology proposed in this study serves as a practical tool for solving complex fleet scheduling problems, which can be adjusted and applied to diverse ex-port demand scenarios to improve the efficiency of real-world fleet operation management.

We agree that omitting stochastic factors (e.g., demand volatility, ETA uncertainty) is a limitation. We have explicitly listed incorporating stochastic/robust optimization for Ro-Ro fleet operations as a core future research direction in the Conclusion to address this gap.

Reviewer 3 Report

Comments and Suggestions for Authors

1)Authors must explain: why they did not add their paper "Wang, S.; Wu, S. Optimizing the Location of Virtual-Shopping-Experience Stores Based on the Minimum Impact on Urban Traffic. Sustainability 202315, 9988. https://doi.org/10.3390/su15139988" in the reference list. Especially, this paper has the highest  plagiarism ratio.

2) No comparisons have been submitted 

3) the optimization Model Formulation must be explained and the constraints must be justified.

4) Authors must provide a flow chart and/or pseudocode for solving criteria/methodology of the  genetic algorithm 

5) The system complexity was not mentioned throughout the paper

6) Authors must justify the selection criteria of the runtime parameters 

7) Authors must provide scientific evidence to explain the following paragraph in the conclusion sub-section:  "The computational results provide detailed sailing and loading/unloading schedules for each vessel. "  

Author Response

Comment 1: Authors must explain: why they did not add their paper "Wang, S.; Wu, S. Optimizing the Location of Virtual-Shopping-Experience Stores Based on the Minimum Impact on Urban Traffic. Sustainability 2023, 15, 9988. https://doi.org/10.3390/su15139988" in the reference list. Especially, this paper has the highest plagiarism ratio.

Response 1: Thank you for your question regarding the reference. The paper “Wang, S.; Wu, S. Optimizing the Location of Virtual-Shopping-Experience Stores” differs fundamentally from our study: it focuses on “virtual shopping store location optimization” (a urban traffic-related issue), while our work addresses “automakers’ self-owned Ro-Ro fleet scheduling” (a maritime logistics problem). The two studies have distinct research questions, mathematical models, and solution algorithms.

The only overlap is partial common authors, which may lead to minor similarities in writing style—but this does not constitute content relevance. Thus, it was not included in the reference list. We confirm no plagiarism between the two papers and have double-checked our manuscript’s originality.

Comment 2: No comparisons have been submitted.

Response 2: Thank you for your comment on "no comparisons". We understand this may refer to a lack of comparison between our optimization results and actual operational data. The reason is that the case focuses on BYD’s self-owned Ro-Ro fleet, which is planned for full delivery around 2025. As of the study’s completion, the fleet has not been fully put into operation, so no actual operational data (e.g., voyage sequences, cost records) is available for comparison. Historical data comparison is thus currently unfeasible.

We have verified the model’s effectiveness via sensitivity analysis and case validation, and will supplement real-data comparisons once the fleet is in operation.

Comment 3: the optimization Model Formulation must be explained and the constraints must be justified.

Response 3: Thank you for your comment on model formulation and constraints. We have supplemented further explanations in the Model Formulation section:

(1) The objective function (minimizing total operational cost) and each constraint (e.g., capacity, time window, carbon emission) are explicitly interpreted to clarify their role in Ro-Ro fleet scheduling.

(2) For constraints, we have reinforced justifications: common-sense constraints (e.g., “one port call per voyage”) align with industry practice, while key constraints (e.g., demand fulfillment) are supported by automakers’ operational needs, as detailed in Section 3.1 and 3.3.

Comment 4: Authors must provide a flow chart and/or pseudocode for solving criteria/methodology of the genetic algorithm.

Response 4: Thank you for your comment on the genetic algorithm (GA) documentation. We have added a detailed flowchart (Figure 3) in the “Solution Algorithm” section, which clearly illustrates the GA’s solving process.

Comment 5: The system complexity was not mentioned throughout the paper.

Response 5: Thank you for your comment on “system complexity”. We understand this may relate to the rationale for choosing the genetic algorithm to address the problem’s complexity.

Our Ro-Ro fleet scheduling involves NP-hard combinatorial optimization, leading to high system complexity. GA is adopted because it avoids the “curse of dimensionality” of exact algorithms, handles infeasible solutions via repair operators, and maintains solution diversity to prevent local optima—effectively adapting to the problem’s complexity, as detailed in the “Solution Algorithm” section.

Comment 6: Authors must justify the selection criteria of the runtime parameters.

Response 6: Thank you for your comment on runtime parameter selection. We have justified the parameter criteria via sensitivity analysis (Section 5.3), which tests key parameters (population size, crossover probability, mutation probability) using the control variable method. The analysis verifies that the selected parameters balance optimization accuracy (avoiding local optima) and computational efficiency, with detailed results shown in Figure 9.

Comment 7: Authors must provide scientific evidence to explain the following paragraph in the conclusion sub-section: “The computational results provide detailed sailing and loading/unloading schedules for each vessel”.

Response 7: Thank you for your comment on the conclusion’s computational results. We provide scientific evidence for the “detailed sailing and loading/unloading schedules” as follows:

Table 7 in Section 5.2 explicitly lists each vessel’s optimized details: port call selection and sequence, sailing speed for each voyage leg, and loading/unloading quantities at each port. These specific data directly constitute the detailed sailing and loading/unloading schedules for every vessel, supporting the conclusion.

Reviewer 4 Report

Comments and Suggestions for Authors
  1. Introduction

This section requires a comprehensive presentation of global market trends for electric vehicles.

I suggest supplementing the data with regional transportation data. This will improve the clarity of your work.

Include a paragraph describing the environmental impact of logistics.

Consider a paragraph that outlines government policies relevant to your work.

Consider a section on a competitive analysis of the transportation strategies of other electric vehicle manufacturers.

Specify the purpose of the work, the research objective, and the time period covered by the study.

 

  1. Literature Review

A detailed literature review is recommended.

Present an analysis of the differences between electric and conventional vehicle transportation.

Try to clearly highlight the theoretical and practical gaps that emerge from your literature research.

You could summarize the literature analyzed in a table, which would increase clarity.

 

  1. Problem and Model Statement

Better justify the assumptions of your proposed model.

Conduct a sensitivity analysis to test the robustness of the model.

Define the weights of the cost elements in detail. This will increase clarity.

There is a lack of modeling of potential threats. Consider adding this element.

There is a need to present alternative objective functions (e.g., environmental aspects).

Consider visualizing the model as a logic diagram. This will improve understandability.

 

  1. Solution Algorithm

Describe in more detail why you are choosing a specific genetic algorithm.

There is no sensitivity testing of parameters (e.g., population size, mutation rate) – I suggest you add this.

Define the convergence criterion precisely. This may increase repeatability.

Include a flowchart of the algorithm steps.

Include an explanation of error handling mechanisms. This will improve robustness.

 

  1. Case Study

Consider adding a multivariate sensitivity analysis of the selected case. This will enhance validity. Analyze the environmental impact of fleet characteristics.

Include a map of the resulting routes. This will improve the reading experience of your article.

Consider comparing your results with actual historical data. This may improve the accuracy of your results.

Consider analyzing several export strategy scenarios.

Consider presenting computational costs or execution times.

 

  1. Discussion

Conduct a deeper analysis and reflection on the effects of geopolitical risk.

Consider incorporating the impact of indirect costs (e.g., warehousing, insurance) and discuss this.

There is a lack of a risk-benefit analysis of outsourcing strategies.

Consider analyzing reliability indicators.

 

  1. Conclusions

Add theoretical implications - it is recommended to highlight specific practical recommendations for theory.

Add practical implications - it is recommended to highlight specific practical recommendations for logistics practice.

What are the implications of your research?

Display the innovative nature of the study.

Describe a list of future research directions.

Add a paragraph on the social or sustainability aspects of the model.

Author Response

Comment 1: Introduction

(1) This section requires a comprehensive presentation of global market trends for electric vehicles.

(2) I suggest supplementing the data with regional transportation data. This will improve the clarity of your work.

(3) Include a paragraph describing the environmental impact of logistics.

(4) Consider a paragraph that outlines government policies relevant to your work.

(5) Consider a section on a competitive analysis of the transportation strategies of other electric vehicle manufacturers.

(6) Specify the purpose of the work, the research objective, and the time period covered by the study.

Response 1: Thank you for your constructive comments on the Introduction section. We have made targeted revisions in the manuscript (revised contents marked in blue) to address each sub-point, with specific modifications detailed below:

(1) Comprehensive presentation of global EV market trends: We expanded the first paragraph to include detailed global NEV market data (Page 1, Lines 31-38).

(2) Supplementation of regional transportation data: We added a new paragraph with regional NEV trade and transportation-related data (Page 2, Lines 39-60).

(3) Description of logistics’ environmental impact: We added a new paragraph focusing on the environmental impact of Ro-Ro shipping logistics (Page 2, Lines 69-74).

(4) Inclusion of relevant government policies: We added a new paragraph detailing government policies shaping NEV adoption and green logistics (Page 2, Lines 75-84).

(5) Competitive analysis of manufacturers’ transportation strategies: We added a new paragraph analyzing three mainstream transportation strategies (full outsourcing strategy, hybrid strategy, and self-built fleet strategy) adopted by global NEV manufacturers (Page 3-4, Lines 115-159).

(6) Specification of work purpose, research objectives, and study time period: We clarified the study’s purpose and outlined three specific research objectives (Page 5, Lines 179-194). Regarding the “time period covered by the study”, we did not specify an explicit start/end time, and the rationale is as follows: This study focuses on theoretical and methodological research on the universal operational scheduling challenges of automakers’ self-built Ro-Ro fleets, rather than empirical data analysis tied to a specific time window. Its conclusions aim to provide generalizable optimization ideas and decision-making tools for automakers to manage their self-owned Ro-Ro fleets now and in the foreseeable future, so an explicit time frame is not set.

Comment 2: Literature Review

(1) A detailed literature review is recommended.

(2) Present an analysis of the differences between electric and conventional vehicle transportation.

(3) Try to clearly highlight the theoretical and practical gaps that emerge from your literature research.

(4) You could summarize the literature analyzed in a table, which would increase clarity.

Response 2: Thank you for your valuable comments on the Literature Review section. We have revised Sub-points (1)-(3) (blue-marked in the manuscript) and explain the reason for not using a literature summary table for Sub-point (4) as follows.

(1) Detailed literature review: We enhanced the review’s depth by: refining “shipper-owned fleet” studies, enriching “shipping company-focused scheduling” research, supplementing latest Ro-Ro studies (Page 5-7).

(2) NEV vs. ICEV transportation differences: A dedicated paragraph is added to analyze core differences (Page 6, Lines 257-273).

(3) Highlighting theoretical and practical gaps: Three clear gaps are refined including objective mismatch (Traditional models (shipping companies) prioritize profit while automakers highlight “demand fulfillment first, cost minimization second”), fleet trait oversight (Automakers’ high-value, demand-volatile vehicle shipping needs flexible scheduling, rarely studied in bulk cargo research), NEV practical gap (emerging self-owned Ro-Ro fleets for NEVs lack targeted scheduling research) (Page 7-8, Lines 325-337).

(4) Rationale for no literature summary table: We avoided a table for three key reasons. First, the revised review is logically classified (4 sections: shipper-owned fleets, shipping company scheduling, Ro-Ro optimization, NEV-ICEV differences), making a table redundant. Second, core literature value lies in context-dependent details that tables cannot fully convey without oversimplification. Third, a table would add redundant info, distracting from identifying research gaps (the review’s core goal). The revised review ensures clarity and depth without a table. We are happy to adjust further if needed.

Comment 3: Problem and Model Statement

(1) Better justify the assumptions of your proposed model.

(2) Conduct a sensitivity analysis to test the robustness of the model.

(3) Define the weights of the cost elements in detail. This will increase clarity.

(4) There is a lack of modeling of potential threats. Consider adding this element.

(5) There is a need to present alternative objective functions (e.g., environmental aspects).

(6) Consider visualizing the model as a logic diagram. This will improve understandability.

Response 3: Thank you for your constructive comments. We have revised Sub-points (1)-(3) and (6) (blue-marked in the manuscript) and explain the rationales for Sub-points (4) and (5) as follows:

(1) Better justify the assumptions: We supplemented detailed justifications for all 7 assumptions (Page 8-9, Lines 372-402).

(2) Sensitivity analysis for model robustness: A sensitivity analysis is added in the Case Study section to test robustness (Section 5.3, Page 18-22, Lines 593-752).

(3) Define cost element weights in detail: Cost weights are clarified in Section 3.1 (consistent with Ro-Ro industry benchmarks, Page 8, Lines 367-370).

(4) Rationale for not adding independent threat modeling:

Potential threats such as geopolitical risks and route disruptions are integrated into the existing model via dynamic parameter adjustments, which avoids redundant model structure while ensuring practical applicability. For route accessibility risks (e.g., the Red Sea conflict), we adjust the “voyage leg distance” parameter—setting the distance of completely impassable legs to an extremely large value to make the model automatically avoid high-risk routes, or increasing the distance proportionally for legs facing delay risks (as verified in the computation results in Case study, refer to Section 5.2). For trade policy risks (e.g., EU NEV tariffs), we dynamically update “regional demand parameters” to reflect policy-driven changes in export demand, and this adjustment method has been validated by the demand sensitivity analysis. This approach aligns with the model’s practical application needs and prevents overcomplication of the optimization framework.

(5) Rationale for not adding alternative environmental objective functions: Environmental requirements are effectively addressed through adding carbon emission constraint (Equation 16). To avoiding multi-objective function complexity, we do not revise the model’s objective function.

(6) Visualize the model as a logic diagram: A model logic diagram (Figure 2) is added (Section 3.3, Page 11).

We are happy to adjust further if needed.

Comment 4: Solution Algorithm

(1) Describe in more detail why you are choosing a specific genetic algorithm.

(2) There is no sensitivity testing of parameters (e.g., population size, mutation rate) – I suggest you add this.

(3) Define the convergence criterion precisely. This may increase repeatability.

(4) Include a flowchart of the algorithm steps.

(5) Include an explanation of error handling mechanisms. This will improve robustness.

Response 4: Thank you for your valuable comments on the Solution Algorithm section. We have revised Sub-points (1)-(5) in Section 4 and Section 5 (revised contents marked in blue), with specific details as follows:

(1) Detail reasons for choosing the specific genetic algorithm (GA): We clarified the selection rationale in Section 4 (Page 12, Lines 466-471).

(2) Add parameter sensitivity testing: A sensitivity testing for key GA parameters (population size, crossover probability, mutation probability) is included in Section 5.3 (Page 18-20 Lines 594-647).

(3) Precisely define the convergence criterion: The convergence criterion is clearly defined in Step 0 of Section 4 (Page 13, Lines 477-483).

(4) Include algorithm steps flowchart: A flowchart of the GA steps (Figure 3) is added in Section 4 (Page 13).

(5) Explain error handling mechanisms: Error handling for infeasible solutions is detailed in Section 4 (Page 14-15, Lines 526-531).

We are happy to adjust further if needed.

Comment 5: Case Study

(1) Consider adding a multivariate sensitivity analysis of the selected case. This will enhance validity.

(2) Include a map of the resulting routes. This will improve the reading experience of your article.

(3) Consider comparing your results with actual historical data. This may improve the accuracy of your results.

(4) Consider analyzing several export strategy scenarios.

(5) Consider presenting computational costs or execution times.

Response 5: Thank you for your valuable comments on the Case Study section. We have revised Sub-points (1), (3), (5), and (6) in Section 5 (revised contents marked in blue) and explain the rationale for Sub-point (2) and (4) as follows:

(1) Multivariate sensitivity analysis: A sensitivity analysis is conducted in Section 5.3, covering three dimensions including algorithm parameters (population size, crossover/mutation probability), transportation demand (European market fluctuation), and export strategy scenarios (priority markets, port dispatch adjustment), verifying model validity.

(2) Map of resulting routes: An optimized fleet route map (Figure 8) is added in Section 5.2. (Page 18).

(3) Rationale for no comparison with actual historical data: We cannot compare with actual historical data for a practical reason.

The case focuses on BYD’s self-owned Ro-Ro fleet, which is planned to be fully delivered around 2025. As of the study’s completion, the fleet has not been fully put into operation, and no actual operational scheduling data (e.g., voyage sequences, cost records) exists. Thus, historical data comparison is currently unfeasible.

(4) Analysis of export strategy scenarios: Three export strategy scenarios (prioritizing delivery to the European market, prioritizing delivery to the Southeast Asian market, adjusting export share of domestic factories) are analyzed in Section 5.3 (Page 21-22, Lines 697-752).

(5) Computational costs/execution times: The execution time is specified in Section 5.2 (Page 17, Line 576).

We are happy to adjust further if needed.

Comment 6: Discussion

(1) Conduct a deeper analysis and reflection on the effects of geopolitical risk.

(2) Consider incorporating the impact of indirect costs (e.g., warehousing, insurance) and discuss this.

(3) There is a lack of a risk-benefit analysis of outsourcing strategies.

(4) Consider analyzing reliability indicators.

Response 6: Thank you for your insightful comments on the Discussion section. We have revised Sub-points (1), (2), and (3) in Section 5.4 (revised contents marked in blue) and explain the rationale for Sub-point (4) as follows:

(1) Deeper analysis of geopolitical risk effects: We enhanced the analysis by quantifying geopolitical risk integration into the model. For route accessibility risks (e.g., Red Sea conflict), we adjust the “voyage leg distance matrix” (set impassable legs to a large value, or increase distance by delay ratio), validated by the optimization result in the case study (Section 5.2); for trade policy risks (e.g., EU NEV tariffs), we dynamically update “regional demand parameters”, validated by demand sensitivity analysis (Section 5.3). For the details, please refer to Page 23 (Lines 776-800).

(2) Incorporate indirect costs (warehousing, insurance): We added a dedicated discussion on indirect costs including warehousing cost and insurance cost in Section 5.4, and propose integrating these into the cost framework in future models. For the details, please refer to Page 23-24 (Lines 801-814).

(3) Risk-benefit analysis of outsourcing strategies: We compared self-owned and outsourcing strategies in Section 5.4. The benefits of outsourcing include lowering upfront investment, flexible capacity adjustment, and mature third-party port/risk systems; while the risks of outsourcing include cost volatility, capacity unreliability, strategic dependence, highlighting the value of self-owned fleets for stable scheduling. For the details, please refer to Page 23 (Lines 754-775).

(4) Rationale for no additional reliability indicator analysis:

The two core reliability indicators (demand fulfillment, on-time arrival) are embedded as mandatory constraints in the model, making independent analysis redundant. The demand fulfillment requirement is embodied by Constraints 4–5 (Equations 5–6) which ensure loading/unloading quantities match port dispatch/demand. The on-time arrival requirement is embodied by Constraints 6–7 (Equations 7–8) which restrict vessel arrival to port time windows. All optimized routes (Table 7) satisfy these two constraints, so the optimization results inherently guarantee reliability, and additional analysis would not enhance validity.

We are happy to adjust further if needed.

Comment 7: Conclusions

(1) Add theoretical implications - it is recommended to highlight specific practical recommendations for theory.

(2) Add practical implications - it is recommended to highlight specific practical recommendations for logistics practice.

(3) What are the implications of your research?

(4) Display the innovative nature of the study.

(5) Describe a list of future research directions.

(6) Add a paragraph on the social or sustainability aspects of the model.

Response 7: Thank you for your valuable comments on the Conclusion section. We have made targeted revisions in the manuscript (marked in blue), with specific modifications detailed below:

(1)-(4) Integrated theoretical/practical implications & study innovation: We have added a new paragraph to demonstrate the theoretical/practical implications. For the details, please refer to Page 24-25 (Lines 845-862).

(5) List of future research directions: Five specific directions are outlined in the Conclusion. For the details, please refer to Page 25 (Lines 877-895).

(6) Social/sustainability aspects: A dedicated paragraph is added in the Conclusion. For the details, please refer to Page 25 (Lines 863-876).

We are happy to adjust further if needed.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Very good revision, I recommend accepting it immediately.

Author Response

Thank you very much for your positive feedback and kind recommendation. We greatly appreciate the valuable suggestions you provided and your recognition of our revisions.

Reviewer 3 Report

Comments and Suggestions for Authors

The required answers have been revised 

Also, I need to justify the compariosns between the proposed system and any similar or semi simliar 

Author Response

Comment 1: The required answers have been revised. Also, I need to justify the comparisons between the proposed system and any similar or semi similar.

Response 1: Thank you for your valuable comment on justifying comparisons between the proposed system and similar/semi-similar works. To address this, we have supplemented systematic comparative analysis in Section 5.4 of the revised manuscript (highlighted in blue).

First, a comparison table (Table 9) is added to explicitly contrast the proposed model with two representative categories of existing studies: carrier-focused Ro-Ro scheduling models (e.g., Zhen et al., 2025) and cargo owner-focused bulk fleet scheduling models (e.g., Song & Jin, 2025). The comparison covers core dimensions including research perspective, core objective, decision variables, key constraints, and applicable scenarios, clearly delineating the uniqueness of the proposed model.

Second, we elaborate on two fundamental distinctions: (1) The objective function prioritizes automakers’ export demand fulfillment (primary constraint) while minimizing operational costs (secondary goal), differing from carrier models that focus on profit maximization and may sacrifice small-batch demand. (2) The integration of three-dimensional dynamic decisions (port call sequence/selection, loading/unloading quantities, voyage speed) and Ro-Ro-specific constraints (e.g., CEU capacity) addresses the “small-batch, multi-destination” trait of vehicle exports, which bulk fleet models (optimized for fixed-port bulk shipments) cannot accommodate.

We also clarify limitations in numerical comparison. Due to unavailable detailed parameters of peer models and the fact that BYD’s Ro-Ro fleet is not fully operational (no real-world benchmark data), direct quantitative comparison with identical inputs is currently unfeasible. Future research will supplement such comparisons once relevant data and parameters are accessible.

Back to TopTop