Speaker
Description
The rapid development of artificial intelligence has brought transformative opportunities to the field of structural health monitoring (SHM) for civil infrastructure. Recent advancements in vision-language models (VLMs) and large language models (LLMs) have demonstrated remarkable capabilities across various tasks, including image classification, object detection, semantic segmentation, instance segmentation, and question-answering. These technologies enable comprehensive, efficient, and intelligent analysis of structural conditions, facilitating early detection of potential issues in complex infrastructures.
This work explores the integration of these cutting-edge models into SHM workflows, showcasing their ability to process multimodal data (e.g., images, sensor data, and textual descriptions) and provide actionable insights. By leveraging the strengths of VLMs and LLMs, such as natural language understanding and advanced visual feature extraction, we propose novel applications for automated damage detection, anomaly assessment, and real-time monitoring. Preliminary results highlight the potential of these models to enhance decision-making processes and reduce human intervention in infrastructure maintenance. This study will provide an overview of state-of-the-art AI methodologies, discuss their strengths and limitations in the context of SHM, and outline future research directions for applying these technologies to improve the safety and resilience of modern civil infrastructure.
