Ontology-approach for Modelling of Maintenance Capability of Complex System Environments

This paper presents a structured, ontology-driven approach for modeling maintenance capability in complex system environments, with a focus on military contexts. Grounded in the DOTMLPFI capability framework, the authors develop a formal ontology using BFO and IOF Core Ontology, supplemented with SKOS to enable lightweight semantic relationships. The approach is validated using competency questions (CQs) and evaluated through a case study of the Finnish Navy’s Squadron 2020 (SQ2020) Corvette Program.

The ideas of this paper are innovative and interesting. The manuscript is well written, and the contributions are clear. It clearly illustrates the motivation for using ontologies in the maintenance domain, and it effectively integrates systems engineering principles with semantic technologies. The authors demonstrate how formal ontology can address the challenges of heterogeneous data integration, aligning strategic goals with operational requirements. However, I have some minor concerns and suggestions, which are outlined below.

1) In the literature review, to better highlight the novelty of the proposed approach, a more detailed comparison with similar ontology-based models in related domains (e.g., predictive maintenance, asset management ontologies) would be beneficial.

2) The ontology remains at a proof-of-concept stage. Many components (e.g., Organization and Personnel) are inherited without full recontextualization, which may limit the model’s expressiveness and reuse in divergent domains. The authors should elaborate more on current coverage gaps and plans for iterative refinement.

3) The validation primarily uses CQs and logical consistency tests. While methodologically appropriate, the evaluation lacks practical performance metrics or empirical feedback from real-world use (e.g., usability, maintainability, or decision-making support). A more substantial evaluation phase—possibly including user studies or scenario-based benchmarking—would be desirable.

4) The proposed ontology is deeply embedded in military paradigms, and its adaptability to civilian industries is only briefly mentioned. A comparative discussion or pilot in a non-military setting could greatly enhance the paper’s relevance and scope.

Other minor typos:
- Removing ’Predictive‘ in “maintenance types in addition to corrective and predictive are, for example, Predictive, Prescriptive, Condition-based and Scheduled.” (Page 9)
- Please use the full name of “CQs” when it first appears in the manuscript
- The provided Long-term stable URL for resources lacks a README file.

The paper studies an important aspect of the Semantic Web: the application of ontologies in real-world tasks. It first provides a detailed review of modeling issues in maintaining the capability of complex systems environments and then describes how the task was executed within six months by the team of experts. The paper also includes a list of ontological elements and competency questions for testing.
Although the paper presents a concrete application project related to the Semantic Web, I am concerned that its current form may not be suitable for readers of the Semantic Web Journal, who generally expect in-depth technical discussions of modeling choices and ontology design decisions. The current discussions remain high-level, lacking formal definitions, concrete examples, and analyses of the advantages and disadvantages of different approaches. The Materials and Methods section, in particular, needs to be rewritten to help readers understand the specific modeling problems encountered and the solutions proposed.

For example, how exactly is the DOTMLPFI framework employed in constructing the ontology? The authors state that “this research employed a multi-phase ontology engineering methodology to formally represent military maintenance capability within the context of the DOTMLPFI framework,” but readers would expect a clear explanation of how this was achieved. To make the contribution reproducible and instructive, the authors should provide a formal definition of the task and detailed descriptions of the methods used to address it.

Regarding Domain Knowledge Refinement, the paper mentions that only the domains and ranges of object properties were enriched. Are there other refinement carried out? What reasoning tasks are used in the decision-support context?

For Reasoning-Based Verification and Ontology Validation, the authors note that Protégé’s explanation facility (Debugger) was used to analyze and resolve missing inferences or inconsistencies. However, the paper should explicitly describe which missing inferences or inconsistencies were identified and how they were resolved during ontology development.

The claim that “Facility and Material concepts were fully supported through well-defined roles and functions” also requires justification. How can readers assess whether these concepts are indeed well-defined? How do they compare to definitions in existing ontologies?

Similarly, the statement that “Interoperability and Leadership capabilities were effectively captured via logical relationships to systems, processes, and planning elements” is not supported by an explanation of how this was achieved.

In Figure 5, the authors state that “each test case targets a specific logical pattern, such as disjoint classes, inverse property inference, cardinality constraints, or disjointness conditions.” However, the paper does not show where these logical patterns can be found or how they were tested.

The authors claim that the resulting ontology is aligned with foundational ontologies, supports reasoning and interoperability, and has been validated against practical knowledge needs. However, the paper does not specify how this evaluation was performed or what entailments were tested. Readers need these details to understand the modelling insights and lessons learned from this work.

The related work section discusses several aspects of modelling maintenance capability but fails to comment on how the current approach differs from or improves upon existing work. The section “Basis for Ontology” seems unnecessary for this journal’s audience, as the background information it contains is already well known to readers.

Figures 1 and 2 are taken from the literature, but the paper does not explain why they are important to be included the present study.

Finally, there are several editorial and formatting issues that make the paper unsuitable for publication in its current form, for example:
* The authors’ affiliations and email addresses are missing.
* The section title “References” is missing.
* Several citations are incomplete (e.g., the authors or links for [18] are missing, as well as the conference/journal details for [26]).
* There are some grammatical and typographical errors (e.g., “software tools or” → “software tools for”).