Structural Health Monitoring and Earthquake/Climate Proofing by means of AI Sensor Fusion, Digital Twins and Smart Oracles

Introduction

The rising intensity of climate and weather‑related extremes, coupled with seismic risks, threatens the resilience of the built environment. This paper presents an integrated framework for Structural Health Monitoring (SHM) and earthquake/climate proofing that unites AI‑driven sensor fusion, Digital Twins (DTs), and Smart Oracles within the blockchain‑enabled decentralized data and knowledge ecosystem developed by the Horizon Europe project BUILDCHAIN.

Methods

Heterogeneous structural, environmental, climatic, and meteorological data streams are merged into a unified analytical pipeline, where AI models process vibration, acceleration, temperature, humidity, and related signals to extract modal properties, detect anomalies, and update structural performance indicators. Trustworthy data flows are ensured through a Decentralized Oracle Network (DON) connected to InterPlanetary File System (IPFS)-based distributed storage. Climate and weather datasets—historical, real‑time, and forecasted—are transformed into Knowledge Assets (KAs), integrated into the Decentralized Knowledge Graph (DKG), validated through a reputation‑based oracle mechanism, and stored immutably on the IPFS. AI‑based fusion algorithms reconcile sensor observations with physics‑based (i.e., Finite Element) models, triggering automated updating routines that refine key structural parameters (i.e., masonry or timber elastic moduli), which critically influence predictive accuracy under seismic and climate‑induced loads.

Results

Pilot applications involving real buildings demonstrate the framework’s ability to reduce discrepancies between as‑designed and as‑built performance. In two tall cross‑laminated timber (CLT) buildings, AI‑driven fusion and DT‑based updating refined elastic modulus scaling factors derived from modal measurements, improving vibration‑serviceability predictions. Deployments in the Hospital Real of Granada and the Palazzo Poniatowski Guadagni validated the approach with open, scalable SHM architectures, enabling quasi‑real‑time operational modal analysis and dynamic DT generation.

Conclusions

Combining AI‑driven sensor fusion, DTs and Smart Oracles with decentralized data governance enhances SHM reliability, improves predictive performance, and supports real‑time assessment of structural condition and vulnerability under seismic and climate stressors.