Semantic Interoperability of Multi-Agent Systems in Autonomous Maritime Domains

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript entitled Semantic Interoperability of Multi-Agent Systems in Autonomous Maritime Domains addresses a highly relevant and timely topic in the context of the digital transformation of the maritime sector. The paper is clearly written, well structured, and presents an interesting contribution by proposing an ontology-based framework to support semantic interoperability, with a concrete application to fire protection systems onboard autonomous ships. Below I provide specific comments and suggestions organized by section:
For the introduction, it provides a comprehensive background and highlights the importance of semantic interoperability in multi-agent maritime systems. The references are appropriate and up-to-date. I recommend slightly clarifying the distinction between existing initiatives (e.g., e-Navigation, AUTOSHIP) and the novel contribution of this paper. In its current form, the gap is stated, but the added value of the proposed framework compared to those initiatives could be emphasised more strongly.
For the second section, it is well structured and informative. The functional decomposition of ships and ports is clearly explained, and Figures 1 and 2 are helpful for understanding. The discussion on the MASS framework and smart ports is valuable. However, when introducing the potential of multi-agent systems, it would be useful to briefly mention interoperability challenges in actual multi-agent implementations (e.g., message understanding, synchronization, ontology alignment), as these are directly related to the paper’s core topic.
For the conclusions, they are generally aligned with the content of the paper but should be more balanced. The current text suggests that the framework is ready to enable broad multi-agent interoperability across the maritime domain, but the experimental validation remains limited to a specific use case. It is recommended to moderate the claims and explicitly state the current limitations and future validation needs

Author Response

We would like to thank the reviewer for the careful reading of the manuscript and for the constructive comments, which helped us improve the quality and clarity of the paper. Our point-by-point responses are provided below.

COMMENTS 1: I recommend slightly clarifying the distinction between existing initiatives (e.g., e-Navigation, AUTOSHIP) and the novel contribution of this paper. In its current form, the gap is stated, but the added value of the proposed framework compared to those initiatives could be emphasised more strongly.

RESPONSE 1: We have revised the manuscript to more clearly articulate the distinction between the goals and scope of existing initiatives (e-Navigation, AUTOSHIP, S-100) and the specific contribution of our proposed framework. A new paragraph has been added at the end of the relevant subsection to highlight how our ontology-based approach uniquely addresses semantic interoperability in multi-agent coordination, which remains insufficiently explored in current maritime initiatives. 

COMMENTS 2: When introducing the potential of multi-agent systems, it would be useful to briefly mention interoperability challenges in actual multi-agent implementations (e.g., message understanding, synchronization, ontology alignment), as these are directly related to the paper’s core topic.

RESPONCE 2: We have revised the manuscript by adding a short paragraph that explicitly identifies key interoperability challenges in real-world multi-agent implementations, such as message interpretation, temporal synchronization, and ontology alignment. These issues are closely related to our proposed semantic interoperability framework and help contextualize the need for such an approach. The addition is placed in the introductory part of Section 2 to ensure a coherent transition from general multi-agent concepts to the semantic interoperability problem our work addresses.

COMMENTS 3: For the conclusions, they are generally aligned with the content of the paper but should be more balanced. The current text suggests that the framework is ready to enable broad multi-agent interoperability across the maritime domain, but the experimental validation remains limited to a specific use case. It is recommended to moderate the claims and explicitly state the current limitations and future validation needs.

RESPONCE 3: The conclusion section has been revised to present a more balanced and realistic reflection of the current state of the framework. Specifically, we now explicitly acknowledge that the experimental validation is limited to a fire protection scenario and that broader validation across multiple operational contexts and stakeholders remains a goal for future work. 

Reviewer 2 Report

Comments and Suggestions for Authors

1. The experimental validation in the paper is simulation-based, centered on a multi-agent fire suppression scenario. While the large number of test cases (650,000) is commendable, the lack of real-world deployment limits the assessment of practical feasibility. The authors are encouraged to consider, at least as future work, the integration of the framework with real maritime systems or testbeds to evaluate its performance under physical constraints and communication uncertainties.
2. The reasoning mechanism presented is primarily rule-based and appears to rely on simple IF–THEN constructs. However, in real-world scenarios, multiple rules may be triggered simultaneously or may produce conflicting actions. The manuscript does not discuss how such conflicts are detected or resolved. It would strengthen the framework if the authors could describe any conflict arbitration mechanisms, such as priority levels, context-aware reasoning, or rule weighting strategies.
3. While the proposed ontology-driven framework is well-structured, the paper does not address how the ontology would be maintained or updated over time. In practical multi-agent systems, new sensors, actuators, or control rules may be introduced, requiring ontology evolution. The authors are encouraged to clarify whether their framework supports dynamic ontology extension, version control, or alignment among distributed agents to ensure semantic consistency during runtime.
4. While the related work section references relevant research, the paper does not provide a structured comparison between the proposed approach and existing semantic interoperability frameworks or multi-agent coordination models. Including a comparative analysis — either in tabular form or through performance benchmarks — would help highlight the novelty and advantages of the proposed method more convincingly.
5. The manuscript demonstrates a solid amount of work in terms of theoretical framework development, ontology modeling, and multi-agent system structuring. The use of RDF/OWL and a layered semantic architecture shows strong conceptual effort. However, the system remains largely at the simulation level, with limited real-world implementation or deployment details. Further development toward integration with physical systems and more advanced reasoning mechanisms would strengthen the work and justify its engineering impact.
6. While the figures generally convey the intended architecture and semantic model, several images suffer from low resolution, small font sizes, and lack of legends or detailed labeling. Some diagrams appear to be screenshots rather than vector-based illustrations. The authors are encouraged to revise the figures using professional tools, improve visual clarity, and ensure consistent styling to meet the standards of academic publication.

Author Response

We would like to thank the reviewer for the careful reading of the manuscript and for the constructive comments, which helped us improve the quality and clarity of the paper. Our point-by-point responses are provided below.

COMMENTS 1: The experimental validation in the paper is simulation-based, centered on a multi-agent fire suppression scenario. While the large number of test cases (650,000) is commendable, the lack of real-world deployment limits the assessment of practical feasibility. The authors are encouraged to consider, at least as future work, the integration of the framework with real maritime systems or testbeds to evaluate its performance under physical constraints and communication uncertainties.

RESPONSE 1: We fully agree that real-world deployment represents a necessary step to assess the practical feasibility and robustness of the proposed framework. Although our extensive simulation campaign provides a valuable initial validation, it cannot capture all the constraints and uncertainties present in physical maritime environments. To address this concern, we have revised the conclusion section to explicitly state this limitation and to outline future research directions involving integration with maritime testbeds and live operational systems. This will enable comprehensive evaluation of the framework under realistic conditions, including sensor variability, communication latency, and system-level disturbances.

COMMENTS 2: The reasoning mechanism presented is primarily rule-based and appears to rely on simple IF–THEN constructs. However, in real-world scenarios, multiple rules may be triggered simultaneously or may produce conflicting actions. The manuscript does not discuss how such conflicts are detected or resolved. It would strengthen the framework if the authors could describe any conflict arbitration mechanisms, such as priority levels, context-aware reasoning, or rule weighting strategies.

RESPONSE 2: We appreciate the reviewer’s insightful observation. Indeed, conflict resolution among simultaneously triggered rules is a critical aspect in practical multi-agent and rule based systems. To address this, we have expanded the section discussing inference rules to include a paragraph on conflict arbitration mechanisms. Specifically, we now explain how rule priorities and contextual conditions are used to manage potential conflicts and ensure consistent system behavior. This addition improves the robustness of the framework and clarifies its applicability in complex, real time maritime environments.

COMMENTS 3: While the proposed ontology-driven framework is well-structured, the paper does not address how the ontology would be maintained or updated over time. In practical multi-agent systems, new sensors, actuators, or control rules may be introduced, requiring ontology evolution. The authors are encouraged to clarify whether their framework supports dynamic ontology extension, version control, or alignment among distributed agents to ensure semantic consistency during runtime.

RESPONSE 3: We have added a dedicated paragraph in the Section 3. (Framework for Semantic Interoperability) to clarify how our approach supports ontology evolution. 

COMMENTS 4: While the related work section references relevant research, the paper does not provide a structured comparison between the proposed approach and existing semantic interoperability frameworks or multi-agent coordination models. Including a comparative analysis — either in tabular form or through performance benchmarks — would help highlight the novelty and advantages of the proposed method more convincingly.

RESPONSE 4: We have revised the manuscript by adding a short paragraph that explicitly identifies key interoperability challenges in real-world multi-agent implementations, such as message interpretation, temporal synchronization, and ontology alignment. These issues are closely related to our proposed semantic interoperability framework and help contextualize the need for such an approach. Also, we have revised the manuscript to more clearly articulate the distinction between the goals and scope of existing initiatives (e-Navigation, AUTOSHIP, S-100) and the specific contribution of our proposed framework. A new paragraph has been added at the end of the relevant subsection to highlight how our ontology-based approach uniquely addresses semantic interoperability in multi-agent coordination, which remains insufficiently explored in current maritime initiatives. 

COMMENTS 5: The manuscript demonstrates a solid amount of work in terms of theoretical framework development, ontology modeling, and multi-agent system structuring. The use of RDF/OWL and a layered semantic architecture shows strong conceptual effort. However, the system remains largely at the simulation level, with limited real-world implementation or deployment details. Further development toward integration with physical systems and more advanced reasoning mechanisms would strengthen the work and justify its engineering impact.

RESPONSE 5: The conclusion section has been revised to present a more balanced and realistic reflection of the current state of the framework. Specifically, we now explicitly acknowledge that the experimental validation is limited to a fire protection scenario and that broader validation across multiple operational contexts and stakeholders remains a goal for future work. 

COMMENTS 6: While the figures generally convey the intended architecture and semantic model, several images suffer from low resolution, small font sizes, and lack of legends or detailed labeling. Some diagrams appear to be screenshots rather than vector-based illustrations. The authors are encouraged to revise the figures using professional tools, improve visual clarity, and ensure consistent styling to meet the standards of academic publication.

RESPONSE 6:
We thank the reviewer for the constructive feedback regarding the figures. In response, all figures have been created using vector-based tools to ensure high resolution and clarity. 

Reviewer 3 Report

Comments and Suggestions for Authors

The paper investigates semantic interoperability in Multi-Agent Systems (MAS) for autonomous maritime domains, proposing an ontology-based framework and validating it through a case study on shipboard firefighting systems. The research topic demonstrates cutting-edge relevance and practical significance, aligning with trends in intelligent shipping. The paper is structurally sound but requires revisions to meet publication standards:

1, While the ontology framework is logically sound, does it significantly surpass existing research (e.g., e-Navigation, S-100)? It is recommended to strengthen comparative analysis with established standards (e.g., S-100), explicitly highlighting its advantages in dynamic data interactions (e.g., real-time fire response scenarios).

2, Does the firefighting case study only verify intra-vessel system interoperability and fails to demonstrate multi-agent collaboration across heterogeneous entities (e.g., vessel-port coordination)? Supplemental experiments validating of MAS cooperation should be added.

3, The description of ontology inference rules is overly generic. Technical details must be refined, including inference rules, security mechanisms, and quantitative analysis of real-time performance (e.g., response latency under operational loads).

Author Response

We would like to thank the reviewer for the careful reading of the manuscript and for the constructive comments, which helped us improve the quality and clarity of the paper. Our point-by-point responses are provided below.

COMMENTS 1: While the ontology framework is logically sound, does it significantly surpass existing research (e.g., e-Navigation, S-100)? It is recommended to strengthen comparative analysis with established standards (e.g., S-100), explicitly highlighting its advantages in dynamic data interactions (e.g., real-time fire response scenarios).

RESPONSE 1: We have revised the manuscript to more clearly articulate the distinction between the goals and scope of existing initiatives (e-Navigation, AUTOSHIP, S-100) and the specific contribution of our proposed framework. A new paragraph has been added at the end of the relevant subsection to highlight how our ontology-based approach uniquely addresses semantic interoperability in multi-agent coordination, which remains insufficiently explored in current maritime initiatives. 

COMMENTS 2: Does the firefighting case study only verify intra-vessel system interoperability and fails to demonstrate multi-agent collaboration across heterogeneous entities (e.g., vessel-port coordination)? Supplemental experiments validating of MAS cooperation should be added.

RESPONSE 2: The conclusion section has been revised to present a more balanced and realistic reflection of the current state of the framework. Specifically, we now explicitly acknowledge that the experimental validation is limited to a fire protection scenario and that broader validation across multiple operational contexts and stakeholders remains a goal for future work. 

COMMENTS 3: The description of ontology inference rules is overly generic. Technical details must be refined, including inference rules, security mechanisms, and quantitative analysis of real-time performance (e.g., response latency under operational loads).

RESPONSE 3: We have enhanced the manuscript in three key areas. First, we have provided a more precise description of an ontology based inference rules. Second, we now explicitly describe the security mechanisms integrated within the agent communication and decision layers, such as integrity checks and role-based data access. Finally, we have added a brief quantitative summary of real-time performance derived from our simulation campaign, including average and worst-case response times under various operational loads. 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

ok

Reviewer 3 Report

Comments and Suggestions for Authors

I agree to accept this manuscript.