Digital Twin and AI Models for Infrastructure Resilience: A Systematic Knowledge Mapping

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This article focuses on “DT + AI,” organizing their application and development for infrastructure. At present, many existing infrastructure systems face the challenge of dynamic updating and require intelligent methods to address it, which has led to extensive research on digital twins and AI in the infrastructure domain. This is a very meaningful review topic that can inspire subsequent development frameworks. However, some important content needs to be supplemented:

1. The review contains substantial analysis related to the topic but does not clearly explain the current DT + AI framework, nor how the two are integrated and applied along a specific pathway.

2. The discussion section is relatively weak and does not sufficiently explore issues and insights regarding future DT + AI development frameworks.

3. The paper summarizes several clusters. What insights are brought by the distribution of these clusters?

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

1. A suggested list of the current research status and development trends for digital twins and artificial intelligence models. 

2. In Section 4.2, what is the basis for the division of the six major research clusters? It is recommended to include a relevant explanation. 

3. The analysis of practical application cases across various fields is somewhat general. It is recommended to add 1-2 detailed case studies.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

Review Report

Summary

This paper presents a systematic bibliometric and content-based review of the integration of Digital Twin (DT) and Artificial Intelligence (AI) technologies in the domain of infrastructure resilience (IR). Using a curated dataset of 108 peer-reviewed publications from the Web of Science (2016–2025), the authors employ co-occurrence and overlay visualizations (via VOSviewer) to map six interconnected research clusters.

The study identifies a temporal shift from early sensor- and CPS-focused work toward integrated, sustainability-oriented applications—such as urban digital twins, predictive maintenance, and environmental resilience—and concludes with a call for interoperable, ethically governed, and adaptive DT-AI ecosystems.

Strengths

1. Comprehensive and Rigorous Methodology: The combination of systematic literature review (SLR), PRISMA-based screening, bibliometric mapping, and qualitative content analysis provides both breadth and depth, enhancing reproducibility and scholarly rigor.

2. Clear Thematic Organization: The identification and description of six research clusters offer a structured, intuitive knowledge map that effectively captures the interdisciplinary convergence of DT, AI, and resilience.

3. Timely and Policy-Relevant Insights: The emphasis on sustainability, data governance, and cross-sectoral interoperability aligns with current global priorities in smart infrastructure and urban planning.

4. Strong Scholarly Foundation: The manuscript is well-grounded in established resilience theory (e.g., resilience triangle), DT frameworks (e.g., Gemini Principles), and recent AI advances (e.g., GNNs), demonstrating solid domain awareness.

Weaknesses

1. Limited Scope Due to Single-Database Reliance: Restricting data collection exclusively to Web of Science may introduce selection bias and omit relevant work from Scopus, IEEE Xplore, or domain-specific repositories (e.g., transportation or energy journals not fully indexed in WoS).

2. Lack of Validation of Bibliometric Clusters: While the clusters are intuitively labeled and supported by representative citations, the paper does not detail how cluster coherence or thematic validity was verified (e.g., through inter-coder reliability or topic modeling metrics like silhouette scores).

3. Insufficient Discussion of Implementation Barriers: Although the paper notes challenges like data fragmentation and cybersecurity, it does not deeply analyze real-world constraints—such as institutional inertia, legacy system integration, or regulatory hurdles—that impede DT-AI deployment in public infrastructure.

4. Future Directions Remain High-Level: The proposed roadmap (e.g., “ethical AI frameworks,” “participatory governance”) lacks concrete methodological or technical pathways for achieving these goals.

Questions for Authors

1. Database Coverage: What is the estimated recall of your search strategy? Could relevant studies in civil engineering or urban planning journals (e.g., Journal of Infrastructure Systems, Cities) have been missed by using only WoS?

2. Interpretability of Clusters: How were the thematic labels for the six clusters derived? Was there an iterative process involving multiple reviewers or computational validation (e.g., Latent Dirichlet Allocation)?

3. Practical Generalizability: Given that most cited applications are sector-specific (e.g., water, ports), how might the proposed integrated DT-AI framework be operationalized across interdependent infrastructure networks (e.g., power-water-transport)?

4. Metrics Alignment: The paper notes misalignment between AI optimization objectives and resilience metrics (e.g., recovery time). Do the authors propose any specific composite indicators or evaluation protocols to bridge this gap?

Final Recommendation

Accept with Minor Revisions

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have satisfactorily addressed all the raised concerns. I recommend the manuscript for acceptance.

Reviewer 2 Report

Comments and Suggestions for Authors

I have no comments.