Physics-Informed Multimodal Thermal Intelligence for Trustworthy Structural Health Monitoring of Critical Aerospace Structures
Year: 2026
Type: preprint
Abstract: This perspective proposes a trustworthy structural-health-monitoring architecture for critical aerospace structures in which no measurement source is treated as absolute ground truth. The framework integrates full-field infrared thermography T(x,y,t), point structural sensors, operational-state information, a multiphysics digital twin, and physics-informed artificial intelligence. The central research question is not merely whether an anomaly exists, but where the inconsistency most probably originates: the physical structure, a point sensor, the infrared measurement system, or the digital model itself. Recent aerospace research has demonstrated real-time sensing digital twins capable of distinguishing structural degradation from sensor faults and quantifying sensor reliability, while NASA has demonstrated synchronized passive thermography and acoustic emission for real-time damage-growth assessment. These advances establish strong prior art for individual components of the proposed framework. The opportunity investigated here is therefore narrower: to use a calibrated full-field thermal field as an independent spatial observer, explicitly assess the integrity of the thermographic measurement chain, and cross-validate it against point sensing and physics-based predictions. The paper defines the architecture, competing diagnostic hypotheses, falsifiable research hypotheses, and a staged validation route. It does not claim experimental validation or definitive patent novelty.
Author: Botelho, Romero Augusto
Language: English
Subfield: Civil and Structural Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7224095825