While supervised structural health monitoring (SHM) systems can potentially accurately detect and localize damage in bridge structures, they require labeled datasets containing healthy and damaged states to develop decision-making models. Obtaining supervised data from in situ bridges is often impractical. Unsupervised algorithms address this challenge by leveraging baseline (i.e., “healthy”) data, thereby eliminating the need for damaged datasets to detect anomalies. Conventional SHM techniques rely on sensor data to assess the structural condition; however, accurately detecting damage remains a challenge, especially when dealing with differing structure types and sizes and varying loading conditions. This study introduces an unsupervised damage detection methodology that aims to classify bridge health using structural response data under variable loading by leveraging deep learning. A variational autoencoder (VAE) network was employed to develop a data-driven methodology to identify structural anomalies based on response data collected during tests of an in situ bridge. Collected datasets included strain time histories representing three structural states (i.e., “healthy,” two levels of damage). The model was trained using strain data collected from healthy bridge loading tests and validated on separate healthy trials, while testing included both healthy and damaged scenarios to evaluate the method’s generalization and damage sensitivity. A VAE network was trained to learn underlying patterns of healthy behavior under moving live load. The training helped develop a compressed strain time-history representation using a network encoder and accurately reconstructed that input from a lower-dimension space. Normalized data reconstruction errors were used to define a damage index that permitted quantitative assessment of structural deterioration. In addition, clustering each test dataset corresponding as a function of structural state helped quantify response deviation from the healthy state. Results demonstrated that the proposed approach effectively detected increasing damage levels, thereby distinguishing different structural states from one another. Results highlighted the potential robustness and adaptability of the proposed approach to real-world SHM applications.
ZouriqMFLinzellDGAzamSE.Image-based strain response estimation of in-situ bridge using recurrent neural networks (RNNs). Structures2025; 73: 108436.
2.
FarrarCRWordenK.An introduction to structural health monitoring. Phil Trans R Soc A Math Phys Eng Sci2007; 365(1851): 303–315.
3.
AlomariQALinzellDGAbu ZouriqM F.Response of bare and CFRP-Retrofitted multi-column piers under post-fire-coupled vehicle collision and air blast. Materials2025; 18(7): 1449.
4.
YangYZhangYTanX.Review on vibration-based structural health monitoring techniques and technical codes. Symmetry2021; 13(11): 1998.
5.
OkashaNMFrangopolDMSaydamD, et al. Reliability analysis and damage detection in high-speed naval craft based on structural health monitoring data. Struct Health Monit2010; 10(4): 361–379.
6.
Abu ZouriqMFLinzellDGNasimiR.Vision-based structural displacement measurements using deep learning and super-resolution. In: JensenJSSchmidtJWFrangopolDM (eds) Bridge Maintenance, Safety, Management, Digitalization and Sustainability, CRC Press, 2024, pp. 735–741.
HumarJBagchiAXuH.Performance of vibration-based techniques for the identification of structural damage. Struct Health Monit2006; 5(3): 215–241.
9.
DoNTGülM.A time series based damage detection method for obtaining separate mass and stiffness damage features of shear-type structures. Eng Struct2020; 208: 109914.
10.
GhobarahAAbou-ElfathHBiddahA.Response-based damage assessment of structures. Earthq Eng Struct Dyn1999; 28(1): 79–104.
11.
FavarelliEGiorgettiA.Machine learning for automatic processing of modal analysis in damage detection of bridges. IEEE Trans Instrum Meas2020; 70: 1–3.
12.
AlampalliSFuGDillonEW.Signal versus noise in damage detection by experimental modal analysis. J Struct Eng1997; 123(2): 237–245.
13.
WuJLiHYeF, et al. Damage identification of bridge structure based on frequency domain decomposition and strain mode. J Vibroeng2019; 21(8): 2096–2105.
14.
NguyenVHMahowaldJMaasS, et al. Use of time-and frequency-domain approaches for damage detection in civil engineering structures. Shock Vib2014; 2014(1): 872492.
15.
KostićBGülM.Vibration-based damage detection of bridges under varying temperature effects using time-series analysis and artificial neural networks. J Bridge Eng2017; 22(10): 04017065.
16.
FarahaniRVPenumaduD.Damage identification of a full-scale five-girder bridge using time-series analysis of vibration data. Eng Struct2016; 115: 129–139.
17.
ChaYJAliRLewisJ, et al. Deep learning-based structural health monitoring. Autom Constr2024; 161: 105328.
18.
SakellariouJSFassoisSDSakarisCS.IWSHM 2017: vibration-based damage localization and estimation via the stochastic functional model–based method: an overview. Struct Health Monit2018; 17(6): 1335–1348.
19.
MoallemiABurrelloABrunelliD, et al. Model-based vs. data-driven approaches for anomaly detection in structural health monitoring: a case study. In: IEEE Int Instrum Meas Technol Conf (I2MTC), Glasgow, UK, 17–20 May 2021, pp. 1–6, New York, NY, USA: IEEE.
20.
WanSGuanSTangY.Advancing bridge structural health monitoring: Insights into knowledge-driven and data-driven approaches. J Data Sci Intell Syst2024; 2(3): 129–140.
21.
BiliszczukJHawryszkówPTeichgraeberM.SHM system and a FEM model-based force analysis assessment in stay cables. Sensors2021; 21(6): 1927.
22.
ZhangZSunC.Structural damage identification via physics-guided machine learning: a methodology integrating pattern recognition with finite element model updating. Struct Health Monit2021; 20(4): 1675–1688.
23.
ZacharakisIGiagopoulosD.Vibration-based damage detection using finite element modeling and the metaheuristic particle swarm optimization algorithm. Sensors2022; 22(14): 5079.
24.
EbrahimianHAstrozaRConteJP, et al. Nonlinear finite element model updating for damage identification of civil structures using batch Bayesian estimation. Mech Syst Signal Process2017; 84: 194–222.
25.
HaidarpourAKongFT.Finite element model updating for structural health monitoring. Struct Durab Health Monit2020; 14(1): 1.
EltounyKGomaaMLiangX.Unsupervised learning methods for data-driven vibration-based structural health monitoring: a review. Sensors2023; 23(6): 3290.
28.
García-MacíasEUbertiniF. Integrated SHM systems: damage detection through unsupervised learning and data fusion. Struct Health Monit Data Sci Tech2021: 247–268. Structural Integrity, vol 21.
29.
De LautourOROmenzetterP. Prediction of seismic-induced structural damage using artificial neural networks. Eng Struct2009; 31(2): 600–606.
30.
ChongJWKimYChonKH.Nonlinear multiclass support vector machine–based health monitoring system for buildings employing magnetorheological dampers. J Intell Mater Syst Struct2014; 25(12): 1456–1468.
31.
DuranBEmoryDAzamYE, et al. A novel CNN architecture for robust structural damage identification via strain measurements and its validation via full-scale experiments. Measurement2025; 239: 115393.
32.
LiangX.Image-based post-disaster inspection of reinforced concrete bridge systems using deep learning with Bayesian optimization. Comput Aided Civ Infrastruct Eng2019; 34(5): 415–430.
33.
YuanFGZargarSAChenQ, et al. Machine learning for structural health monitoring: challenges and opportunities. Sens Smart Struct Technol Civ Mech Aerosp Syst2020; 11379: 1137903.
34.
SartoratoMDe MedeirosRVandepitteD, et al. Computational model for supporting SHM systems design: damage identification via numerical analyses. Mech Syst Signal Process2017; 84: 445–461.
35.
RastinZGhodrati AmiriGDarvishanE.Unsupervised structural damage detection technique based on a deep convolutional autoencoder. Shock Vib2021; 2021(1): 6658575.
36.
ShiSDuDMercanO, et al. A novel unsupervised real-time damage detection method for structural health monitoring using machine learning. Struct Control Health Monit2022; 29(10): e3042.
37.
LiSCaoYGdoutosEE, et al. Intelligent framework for unsupervised damage detection in bridges using deep convolutional autoencoder with wavelet transmissibility pattern spectra. Mech Syst Signal Process2024; 220: 111653.
38.
AkintundeEAzamSERagehA, et al. Unsupervised machine learning for robust bridge damage detection: full-scale experimental validation. Eng Struct2021; 249: 113250.
39.
TibaduizaDAMujicaLERodellarJ, et al. Structural damage detection using principal component analysis and damage indices. J Intell Mater Syst Struct2016; 27(2): 233–248.
40.
KhoaNLZhangBWangY, et al. Robust dimensionality reduction and damage detection approaches in structural health monitoring. Struct Health Monit2014; 13(4): 406–417.
41.
PathirageCSLiJLiL, et al. Structural damage identification based on autoencoder neural networks and deep learning. Eng Struct2018; 172: 13–28.
42.
SilvaMFSantosASantosR, et al. Damage-sensitive feature extraction with stacked autoencoders for unsupervised damage detection. Struct Control Health Monit2021; 28(5): e2714.
43.
SlavicGCampoDBaydounM, et al. Anomaly detection in video data based on probabilistic latent space models. In: IEEE conference on evolving and adaptive intelligent systems (EAIS), Bari, Italy, 27 May 2020, pp. 1–8. New York, NY: IEEE.
44.
NetoMSFinottiRBarbosaF, et al. Structural damage identification using autoencoders: a comparative study. Buildings2024; 14(7): 2014.
45.
MaXLinYNieZ, et al. Structural damage identification based on unsupervised feature-extraction via variational auto-encoder. Measurement2020; 160: 107811.
46.
LinJMaH.Structural damage detection based on the correlation of variational autoencoder neural networks using limited sensors. Sensors2024; 24(8): 2616.
47.
PollastroATestaGBilottaA, et al. Semi-supervised detection of structural damage using variational autoencoder and a one-class support vector machine. IEEE Access; 11: 67098–67112.
48.
ChoiHKimMLeeG, et al. Unsupervised learning approach for network intrusion detection system using autoencoders. J Supercomput2019; 75: 5597–5621.
49.
LiuTWangJLiuQ, et al. High-ratio lossy compression: exploring the autoencoder to compress scientific data. IEEE Trans Big Data2021; 9(1): 22–36.
50.
GonzalezFJBalajewiczM.Deep convolutional recurrent autoencoders for learning low-dimensional feature dynamics of fluid systems. arXiv:1808.01346, 2018.
51.
WangWHuangYWangY, et al. Generalized autoencoder: a neural network framework for dimensionality reduction. In: Proceedings of the IEEE conference on computer vision and pattern recognition workshops, Columbus, OH, 23–28 June, 2014, pp. 490–497. New York, NY: IEEE.
52.
ChenZYeoCKLeeBS, et al. Autoencoder-based network anomaly detection. In: 2018 Wireless telecommunications symposium (WTS), Phoenix, AZ, 17 April 2018, pp. 1–5. New York, NY: IEEE.
53.
ZhouCPaffenrothRC.Anomaly detection with robust deep autoencoders. In: Proceedings of the 23rd ACM SIGKDD international conference on knowledge discovery and data mining, New York, NY, 4 August 2017, pp. 665–674. New York, NY: Association for Computing Machinery (ACM).
54.
AnJChoS.Variational autoencoder based anomaly detection using reconstruction probability. Special Lecture IE2015; 2(1): 1–8.
55.
DongCXueTWangC. The feature representation ability of variational autoencoder. In: 2018 IEEE third international conference on data science in cyberspace (DSC), Guanghzou, China, 18 June 2018, pp. 680–684. New York, NY: IEEE.
56.
JaiswalGRaniRMangotraH, et al. Integration of hyperspectral imaging and autoencoders: benefits, applications, hyperparameter tunning and challenges. Comput Sci Rev2023; 50: 100584.
57.
SetitraMAFanMBensalemZE.An efficient approach to detect distributed denial of service attacks for software defined internet of things combining autoencoder and extreme gradient boosting with feature selection and hyperparameter tuning optimization. Trans Emerg Telecommun Technol2023; 34(9): e4827.
58.
MujicaLERodellarJFernandezA, et al. Q-statistic and T2-statistic PCA-based measures for damage assessment in structures. Struct Health Monit2011; 10(5): 539–553.
