Advancing bridge structural health monitoring using artificial intelligence
Session
Civil Engineering, Infrastructure and Environment
Description
Continuous structural health monitoring (SHM) systems are essential for the effective management of bridges, enabling real-time assessment of structural conditions and early damage detection. While these systems have proven effective, handling the vast amount of data generated over time presents challenges. Recent advancements in Artificial Intelligence (AI) and Machine Learning (ML) offer solutions to enhance damage detection, classification, and condition assessment. However, classifying damage remains difficult due to the limited availability of labeled data from damaged bridges. To address this, Transfer Learning (TL), a subset of ML, offers a promising approach. TL allows models trained on large datasets from different domains, such as other bridges or simulation data, to be adapted for specific bridge monitoring tasks. This technique helps overcome data scarcity and improves the accuracy of damage classification across various bridge structures. Ongoing research demonstrates how ML and TL address key challenges, leading to more effective and reliable bridge health monitoring, ultimately ensuring safety and longevity.
Keywords:
structural health monitoring, machine learning, transfer learning, bridges
Proceedings Editor
Edmond Hajrizi
ISBN
978-9951-982-15-3
Location
UBT Kampus, Lipjan
Start Date
25-10-2024 9:00 AM
End Date
27-10-2024 6:00 PM
DOI
10.33107/ubt-ic.2024.306
Recommended Citation
Venanzi, Ilaria, "Advancing bridge structural health monitoring using artificial intelligence" (2024). UBT International Conference. 6.
https://knowledgecenter.ubt-uni.net/conference/2024UBTIC/CEIE/6
Advancing bridge structural health monitoring using artificial intelligence
UBT Kampus, Lipjan
Continuous structural health monitoring (SHM) systems are essential for the effective management of bridges, enabling real-time assessment of structural conditions and early damage detection. While these systems have proven effective, handling the vast amount of data generated over time presents challenges. Recent advancements in Artificial Intelligence (AI) and Machine Learning (ML) offer solutions to enhance damage detection, classification, and condition assessment. However, classifying damage remains difficult due to the limited availability of labeled data from damaged bridges. To address this, Transfer Learning (TL), a subset of ML, offers a promising approach. TL allows models trained on large datasets from different domains, such as other bridges or simulation data, to be adapted for specific bridge monitoring tasks. This technique helps overcome data scarcity and improves the accuracy of damage classification across various bridge structures. Ongoing research demonstrates how ML and TL address key challenges, leading to more effective and reliable bridge health monitoring, ultimately ensuring safety and longevity.