Climate Vulnerability Assessment of Infrastructure Using Edge‑AI Integrated IoT Systems: A Survey

Approved

1. Relevance and Originality

The paper addresses a highly relevant and urgent research area by focusing on climate vulnerability assessment of critical infrastructure using edge AI integrated IoT systems. The topic aligns well with current global concerns related to climate resilience and adaptive infrastructure planning. The integration of climate risk theory with practical edge computing architectures demonstrates originality, particularly through the linkage of hazard exposure vulnerability constructs with real time sensing and near edge analytics. As a survey, the work offers a comprehensive synthesis rather than proposing a novel algorithm, which is appropriate for its stated scope.

2. Methodology

The paper adopts a structured survey methodology, clearly defining inclusion criteria, technological boundaries, and temporal scope. The classification of sensing modalities, communication protocols, edge AI techniques, and digital twin integration is systematic and well organized. However, the methodology could be strengthened by briefly explaining how literature quality was assessed or prioritized beyond peer review status and adoption maturity. A short description of the search strategy or databases consulted would improve methodological transparency.

3. Validity and Reliability

The paper demonstrates strong reliability through extensive citation of authoritative sources, including IPCC reports, NIST frameworks, and peer reviewed journals. The alignment with established standards such as OGC SensorThings, MQTT, and NIST AI RMF enhances the credibility of the analysis. The discussion of uncertainty calibration, concept drift, and governance further reinforces validity. One area for improvement would be a clearer differentiation between empirically validated deployments and conceptual or early stage implementations discussed in the survey.

4. Clarity and Structure

The manuscript is well structured and logically sequenced, moving from conceptual foundations to system architecture, methods, applications, and open challenges. Figures and tables are effectively used to summarize architectures, workflows, and deployment characteristics. Some sections are information dense, particularly those covering standards and evaluation metrics, but the inclusion of takeaways and notes helps maintain clarity. Minor condensation of repeated explanations could further enhance readability.

5. Results and Analysis

As a survey paper, the analysis focuses on synthesizing reported outcomes rather than presenting new experimental results. This objective is achieved effectively, with clear articulation of performance gains such as reduced latency, improved resilience under connectivity loss, and better data stewardship. The discussion of evaluation scaffolds and decision centric metrics is particularly strong and adds practical value. Including a brief comparative summary highlighting trade offs between edge only, hybrid, and cloud dominant approaches could further strengthen the analytical insight.