Deep Learning for Structural Health Monitoring: Data, Algorithms, Applications, Challenges, and Trends
Open accessThe DOI matches its Crossref record; 264 of the 271 references it cites were found in a registry or an index; 38 of the 105 sources its claims cite were checked for support.
Abstract
Environmental effects may lead to cracking, stiffness loss, brace damage, and other damages in bridges, frame structures, buildings, etc. Structural Health Monitoring (SHM) technology could prevent catastrophic events by detecting damage early. In recent years, Deep Learning (DL) has developed rapidly and has been applied to SHM to detect, localize, and evaluate diverse damages through efficient feature extraction. This paper analyzes 337 articles through a systematic literature review to investigate the application of DL for SHM in the operation and maintenance phase of facilities from three perspectives: data, DL algorithms, and applications. Firstly, the data types in SHM and the corresponding collection methods are summarized and analyzed. The most common data types are vibration signals and images, accounting for 80% of the literature studied. Secondly, the popular DL algorithm types and application areas are reviewed, of which CNN accounts for 60%. Then, this article carefully analyzes the specific functions of DL application for SHM based on the facility’s characteristics. The most scrutinized study focused on cracks, accounting for 30 percent of research papers. Finally, challenges and trends in applying DL for SHM are discussed. Among the trends, the Structural Health Monitoring Digital Twin (SHMDT) model framework is suggested in response to the trend of strong coupling between SHM technology and Digital Twin (DT), which can advance the digitalization, visualization, and intelligent management of SHM.
Overview
CoChat Summary
TL;DR
This systematic review analyzes 337 articles on deep learning (DL) for structural health monitoring (SHM), organizing the literature by data, algorithms, and applications. It finds CNN-dominated research, highlights cracks and bridges as major foci, and proposes a Structural Health Monitoring Digital Twin (SHMDT) framework as a future trend.
Key findings
- The review covers 337 articles, including 23 review articles, retrieved from Web of Science for 1 January 2017–31 March 2023.
- Vibration signals and images are the most common data types, together accounting for 80% of the studied literature; CNN accounts for 60% of DL algorithm studies.
- Cracks are the most scrutinized application, accounting for 30% of research papers; bridges receive the highest proportion of application attention.
- Only 15% of the research literature reaches the practical application stage, mainly for bridges; the paper suggests an SHMDT model framework to advance digitalization, visualization, and intelligent SHM management.
Methods
- Systematic literature review using Web of Science with keywords combining SHM, DL, and CNN; 555 journal articles were retrieved, then manually screened to 337 articles, including 23 reviews.
- The review analyzes the selected literature across three perspectives: data types and acquisition methods, DL algorithms, and application objects/functions.
- Bibliometric and keyword co-occurrence analysis (VOSviewer) and summary tables/figures are used, followed by discussion of challenges and trends and proposal of the SHMDT framework.
Limitations
- The paper states that it focuses more on CNNs, is retrieved only through the Web of Science platform, and covers three broad perspectives, so some detailed content is not discussed thoroughly.
- Evident from the text: including CNN as a search keyword may bias the algorithm coverage toward CNN and underrepresent other DL methods.
- Evident from the text: the search time frame is limited to 2017–31 March 2023, so earlier and later literature is excluded.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures






![Figure 7 CNN architecture diagram [160].](https://figures.cochat.ai/paper-figures/e7e0b79440c245eeb0d3b404d656f2f4/deep-learning-for-structural-health-monitoring-data-algorithms-applications-figure-7.jpg)
![Figure 8 RNN architecture diagram [104].](https://figures.cochat.ai/paper-figures/2379c9cc749f427db034e57d438ced6d/deep-learning-for-structural-health-monitoring-data-algorithms-applications-figure-8.jpg)
![Figure 9 Auto-encoder architecture diagram [104].](https://figures.cochat.ai/paper-figures/b3aa174267ab45428966ee98c89db927/deep-learning-for-structural-health-monitoring-data-algorithms-applications-figure-9.jpg)
![Figure 10 GAN architecture diagram [104].](https://figures.cochat.ai/paper-figures/f438d79480b347c1b92eac5c8d04b12c/deep-learning-for-structural-health-monitoring-data-algorithms-applications-figure-10.jpg)







Verification
Claims the paper makes
25 Valid · 29 Not valid · 34 Not checked · 100 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“In addition, DL has also been applied to SHM to identify, locate, and evaluate various damages by extracting advanced damage features from data such as pictures and vibration signals [15,35,36].”
§ Deep Learning for Structural Health Monitoring: Data, Algorithms, Applications, Challenges, and Trends > 1. Introduction
What the sources say
- Not checked — no full text of the sourcecites [15] Hurtado, 2023, Unsupervised learning-based framework for indirect structural health monitoring using adversarial autoencoder, J. Sound Vibr. source
- Supportedcites [35] Hou, L., Chen, H.S., Zhang, G.M., and Wang, X.Y. (2021). Deep Learning-Based Applications for Safety Management in the AEC Industry: A Review. Appl. Sci., 11. source
Section 3.1.1 (Vision-Based Damage Detection; page unavailable) describes DL methods that learn damage features from images to detect, localize, and assess structural defects. Section 3.1.2 (Vibration-Based Damage Detection; page unavailable) describes DL methods that extract features from sensor data to detect and locate damage. Together, these passages support the paraphrase. (LLM, conf=0.92)
- Partly supportedcites [36] Kao, S.P., Chang, Y.C., and Wang, F.L. (2023). Combining the YOLOv4 Deep Learning Model with UAV Imagery Processing Technology in the Extraction and Quantization of Cracks in Bridges. Sensors, 23. sourceExperimental — accuracy not yet measured
The Introduction (page unavailable) says CNNs extract features from training images to detect concrete cracks, and the Abstract describes identifying and measuring bridge cracks from UAV photographs. This supports the image-based crack example, but our copy does not support the claim about vibration signals or establish that this approach evaluates various kinds of damage. (LLM, conf=0.87)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Every source the paper cites for this claim was checked and supports it.
In the paper
“Damage-sensitive feature extraction can be achieved through image processing techniques, ML, and DL [100], where DL can automatically extract features for image classification, object recognition, and semantic segmentation tasks.”
§ Deep Learning for Structural Health Monitoring: Data, Algorithms, Applications, Challenges, and Trends > 3. Data Types > 3.2. Image
What the sources say
- Supportedcites [100] Hamishebahar, Y., Guan, H., So, S.P., and Jo, J. (2022). A Comprehensive Review of Deep Learning-Based Crack Detection Approaches. Appl. Sci., 12. source
Section 2, “Automatic Feature Extraction” (p. 4), says deep architectures extract features themselves without handcrafted feature engineering. The same section (p. 5) groups image-based crack-detection approaches using deep architectures into image classification, object recognition, and semantic segmentation. The review discusses these tasks in the context of structural health monitoring. (LLM, conf=0.92)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“The problem of data temporal and spatial feature extraction can be solved using a combination of the CNN and RNN methods, such as CNN with GRU [64,109,123], CNN with LSTM [27,178,179,180,181,182], CNN with Auto-encoder and [183], CNN with echo state networks [113], etc.”
§ Deep Learning for Structural Health Monitoring: Data, Algorithms, Applications, Challenges, and Trends > 4. Deep Learning Algorithms > 4.1. Convolutional Neural Network
What the sources say
- Not checked — no full text of the sourcecites [64] Li, 2022, A new dam structural response estimation paradigm powered by deep learning and transfer learning techniques, Struct. Health Monit. source
- Not checked — no full text of the sourcecites [109] Yang, 2021, A data-driven structural damage detection framework based on parallel convolutional neural network and bidirectional gated recurrent unit, Inf. Sci. source
- Not checked — no full text of the sourcecites [123] Yang, 2020, A hierarchical deep convolutional neural network and gated recurrent unit framework for structural damage detection, Inf. Sci. source
- Supportedcites [27] Parida, L., Moharana, S., Ferreira, V.M., Giri, S.K., and Ascensao, G. (2022). A Novel CNN-LSTM Hybrid Model for Prediction of Electro-Mechanical Impedance Signal Based Bond Strength Monitoring. Sensors, 22. source
Section 2.8 identifies the paper’s model as a CNN–LSTM hybrid: the CNN extracts features and the LSTM learns sequences. This supports its use as an example of combining CNN and recurrent methods, though the paper describes its application to EMI signals rather than using the phrase “spatial and temporal feature extraction.” (LLM, conf=0.88)
- Not checked — no full text of the sourcecites [178] Dang, 2021, Data-Driven Structural Health Monitoring Using Feature Fusion and Hybrid Deep Learning, IEEE Trans. Autom. Sci. Eng. source
- Not checked — no full text of the sourcecites [179] Ma, 2021, Predicting Mechanical State of High-Speed Railway Elevated Station Track System Using a Hybrid Prediction Model, KSCE J. Civ. Eng. source
- Supportedcites [180] Shin, R., Okada, Y., and Yamamoto, K. (2022). Application of C-LSTM Networks to Automatic Labeling of Vehicle Dynamic Response Data for Bridges. Sensors, 22. source
Section 2.2 describes C-LSTM as a combination of CNN and RNN, with the CNN layer handling spatial relationships. It says the model is designed to learn and extract spatio-temporal features simultaneously from vehicle vibration data, supporting the cited CNN–LSTM example. (LLM, conf=0.90)
- Not checked — no full text of the sourcecites [181] Du, 2022, Response Prediction Based on Temporal and Spatial Deep Learning Model for Intelligent Structural Health Monitoring, IEEE Internet Things J. source
- Not checked — no full text of the sourcecites [182] Yue, 2022, Ultra-high precise Stack-LSTM-CNN model of temperature-induced deflection of a cable-stayed bridge for detecting bridge state driven by monitoring data, Structures source
- Supportedcites [183] Parziale, 2022, Vibration-based structural health monitoring exploiting a combination of convolutional neural networks and autoencoders for temperature effects neutralization, Struct. Control Health Monit. source
Citation [183] is attached to the CNN-with-autoencoder example. Section 2.2 says the CNN extracts damage-related dynamic features and autoencoders are used to neutralize temperature effects; the Conclusion describes the combined algorithm. The citation supports that example, rather than the sentence’s broader characterization of the listed combinations as CNN–RNN methods. (LLM, conf=0.88)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on an assumption not shown here.
Every source the paper cites for this claim was checked and supports it.
In the paper
“Parziale et al. [183] proposed a coupling method of a CNN and autoencoder to neutralize the influence of temperature change and improve the accuracy of damage detection on a limited dataset.”
§ Deep Learning for Structural Health Monitoring: Data, Algorithms, Applications, Challenges, and Trends > 4. Deep Learning Algorithms > 4.1. Convolutional Neural Network
What the sources say
- Supportedcites [183] Parziale, 2022, Vibration-based structural health monitoring exploiting a combination of convolutional neural networks and autoencoders for temperature effects neutralization, Struct. Control Health Monit. source
The paper proposes coupling a CNN with autoencoders to reduce temperature-related errors in damage detection. Section 3.3 reports a lower, more uniform false-alarm rate across the temperatures tested, and the Conclusion says the combined algorithm improves performance while allowing a limited training dataset. The demonstration is a numerical structural-beam case study. (LLM, conf=0.93)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Li et al. [197] proposed a transformer-based time series prediction framework, verified by bridge strain data, and was found to have a more minor error than LSTM.”
§ Deep Learning for Structural Health Monitoring: Data, Algorithms, Applications, Challenges, and Trends > 4. Deep Learning Algorithms > 4.5. Others
What the sources say
- Supportedcites [197] Li, Z.Q., Li, D.S., and Sun, T.S. (2022). A Transformer-Based Bridge Structural Response Prediction Framework. Sensors, 22. source
The Conclusions report a Transformer-based time-series prediction framework verified using six months of concrete-bridge strain data. The Experimental Results state that its mean prediction error was 19.2%–55.5% of the LSTM method’s mean error. (LLM, conf=0.94)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
49 more claims · 6 with their cited sources checked
Create a free account to see all 188 claims and to verify the 4 references still open.
- Probably not a substantial claim. 106
- It may state two separate claims as one. 84
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Deep Learning for Structural Health Monitoring: Data, Algorithms, Applications, Challenges, and Trends > 5. Application Objects and Functions > 5.1. Facility Components
“Nguyen et al. [218] developed a 1D-CNN model to extract and learn the optimal features from the original electromechanical impedance signals to assess the severity of prestress loss in reinforced concrete beams.”
§ Deep Learning for Structural Health Monitoring: Data, Algorithms, Applications, Challenges, and Trends > 5. Application Objects and Functions > 5.1. Facility Components
“Nguyen et al. [218] designed 16 damage states for steel pipes and used a CNN to achieve high-precision detection of different welding types, positions, and severity.”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Deep Learning for Structural Health Monitoring: Data, Algorithms, Applications, Challenges, and Trends > 3. Data Types > 3.4. Guided Wave
“to locate and evaluate damage in aluminum plates [65,149]”
§ Deep Learning for Structural Health Monitoring: Data, Algorithms, Applications, Challenges, and Trends > 5. Application Objects and Functions > 5.2. Facilities
“Li et al. [65] decomposed the time series of dam displacement into seasonal, trend, and residual components, adopted extremely randomized trees and stacked LSTM to predict the displacement of each component, and finally obtained the total displacement of the dam.”
Cited works that were retracted
Works this paper cites that were later retracted, from the retraction notices publishers deposit. Major when the notice came out before this paper was published.
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
“Zhao, 2021, Efficient Detection of Large-Scale Multimedia Network Information Data Anomalies Based on the Rule-Extracting Matrix Algorithm, Adv. Multimed., 2021, 3299891, 10.1155/2021/3299891”
Cited work 10.1155/2021/3299891 · retracted 2023-08-16 · notice 10.1155/2023/9786723 · this paper published 2023-10-30
Citations
Cites
1 retracted
Items 201–264 of 264
Seventekidis, P., & Giagopoulos, D. (2023) · Mechanical Systems and Signal Processing · doi:10.1016/j.ymssp.2022.109741 · Cited by 37
Ye, X., Ma, S., Liu, Z., Ding, Y., Li, Z., & Jin, T. (2022) · Structural Control and Health Monitoring · doi:10.1002/stc.3128 · Cited by 37 · cited 1×
“One of the reasons for this phenomenon is data problems, including less corrupted data and noise effects.”
BackgroundZhang, B., Hong, X., & Liu, Y. (2021) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2021.3091204 · Cited by 36 · cited 1×
“They can be applied for dimensionality reduction [74,75], image generation [43,76], lost data reconstruction [77–79], AE source localization [80,81], crack detection [82–84], and so on.”
BackgroundHan, G., Kim, Y., Kim, H., et al. (2021) · Expert Systems with Applications · doi:10.1016/j.eswa.2021.115863 · Cited by 36
Deng, Y., Ju, H., Li, Y., Hu, Y., & Li, A. (2022) · International Journal of Architectural Heritage · doi:10.1080/15583058.2022.2153234 · Cited by 35
Self-citationEvaluation of ASR in concrete using acoustic emission and deep learning
Ai, L. M., Soltangharaei, V., & Ziehl, P. H. (2021) · Nuclear Engineering and Design · doi:10.1016/j.nucengdes.2021.111328 · Cited by 34 · cited 1×
“The number of neurons in the input and output layers is the same [188], which aims to reconstruct the input data.”
BackgroundA Transformer-Based Bridge Structural Response Prediction Framework
Li, Z., Li, D., & Sun, T. (2022) · Sensors · doi:10.3390/s22083100 · Cited by 34 · cited 1×
“Li et al. [197] proposed a transformer-based time series prediction framework, verified by bridge strain data, and was found to have a more minor error than LSTM.”
ResultPerformance Comparison of Multiple Convolutional Neural Networks for Concrete Defects Classification
Arafin, P., Issa, A., & Billah, A. H. M. M. (2022) · Sensors · doi:10.3390/s22228714 · Cited by 34
Damage identification using the PZT impedance signals and residual learning algorithm
Al-Azzawi, O., & Wang, D. (2021) · Journal of Civil Structural Health Monitoring · doi:10.1007/s13349-021-00505-9 · Cited by 33
Yang, L., & Xu, F. (2020) · IEEE Access · doi:10.1109/access.2020.3012521 · Cited by 32
Generative Adversarial Network for Damage Identification in Civil Structures
Rastin, Z., Ghodrati Amiri, G., & Darvishan, E. (2021) · Shock and Vibration · doi:10.1155/2021/3987835 · Cited by 31
A channel-spatial-temporal attention-based network for vibration-based damage detection
Liao, S., Liu, H., Yang, J., & Ge, Y. (2022) · Information Sciences · doi:10.1016/j.ins.2022.05.042 · Cited by 31 · cited 3×
“Liao et al. [122] proposed a channel-spatial-temporal a tt ention-based network to detect vibration-based damage.”
BackgroundMethodYue, Z., Ding, Y., Zhao, H., & Wang, Z. (2022) · Structures · doi:10.1016/j.istruc.2022.09.011 · Cited by 31
Son, H., Pham, V.-T., Jang, Y., & Kim, S.-E. (2021) · Sensors · doi:10.3390/s21093118 · Cited by 29 · cited 1×
“…CapsNets to localize and quantify the stiffness reduction damage in beams, showing that the method achieved better results than CNNs. Son et al. [196] proposed that a Graph Neural Network (GNN) can learn graph structure data, which overcame the limitation that a CNN can only learn grid…”
ResultChen, J., Wu, W., Ren, Y., & Yuan, S. (2021) · Sensors · doi:10.3390/s22010307 · Cited by 29 · cited 2×
“Chen et al. [21] used two PZT sensors to collect the GW signal of the original state of the aluminum plate and the GW signal during the gradual generation of fatigue cracks.”
MethodBolt Loosening Detection Using Key-Point Detection Enhanced by Synthetic Datasets
Lu, Q., Jing, Y., & Zhao, X. (2023) · Applied Sciences · doi:10.3390/app13032020 · Cited by 28 · cited 3×
“…defect images of bridges, frame structures, dams, tunnels, and other infrastructures for crack detection [40–44], bolt loosening detection [45–47], rebar surface defect detection [48], delamination and reinforcement exposure detection [49], displacement detection [50–53], voids detection…”
BackgroundMethodDeep learning visual interpretation of structural damage images
Gao, Y., & Mosalam, K. M. (2022) · Journal of Building Engineering · doi:10.1016/j.jobe.2022.105144 · Cited by 27 · cited 2×
“Ye et al. [222] constructed an image dataset called the bridge crack library, which collected 11,000 high-pixel labeled images of 50 Bridges in 2 years and was verified using a DNN for crack detection.”
MethodTransfer Learning for CNN-Based Damage Detection in Civil Structures with Insufficient Data
Chamangard, M., Amiri, G. G., Darvishan, E., & Rastin, Z. (2022) · Shock and Vibration · doi:10.1155/2022/3635116 · Cited by 27
An Autoencoder-Based Deep Learning Approach for Load Identification in Structural Dynamics
Rosafalco, L., Manzoni, A., Mariani, S., & Corigliano, A. (2021) · Sensors · doi:10.3390/s21124207 · Cited by 26
An Automatic Error Detection Method for Machine Translation Results via Deep Learning
Zhang, W. (2023) · IEEE Access · doi:10.1109/access.2023.3280549 · Cited by 26
Magnetic Particle Inspection Optimization Solution within the Frame of NDT 4.0
Sacarea, A. I., Oancea, G., & Pârv, L. (2021) · Processes · doi:10.3390/pr9061067 · Cited by 25
Baquerizo, J., Tutivén, C., Puruncajas, B., Vidal, Y., & Sampietro, J. (2022) · Mathematics · doi:10.3390/math10071131 · Cited by 25 · cited 1×
“The typical ratios of the training set, verification set, and test set are 6:2:2 [17,110], 7:2:1 [22,44], 8:1:1 [128,199] and so on.”
BackgroundBarraza, J. F., Droguett, E. L., Naranjo, V. M., & Martins, M. R. (2020) · Applied Soft Computing · doi:10.1016/j.asoc.2020.106732 · Cited by 24
Wave based damage detection in solid structures using spatially asymmetric encoder–decoder network
Wuttke, F., Lyu, H., Sattari, A. S., & Rizvi, Z. H. (2021) · Scientific Reports · doi:10.1038/s41598-021-00326-2 · Cited by 24 · cited 1×
“…basis, One-Dimensional Convolutional Neural Networks (1D-CNN) with simple architecture and low computational complexity are applied to SHM to directly process 1D data for crack detection [56,57], corrosion detection [58], multi-type damage identification [1], abnormal data detection [59,60], etc.”
Ju, H., Deng, Y., Zhai, W., & Li, A. (2022) · Sensors and Materials · doi:10.18494/sam4000 · Cited by 24 · cited 2×
“…and Gated Recurrent Unit (GRU), it can not only realize multi-type damage identification and location [61], crack detection [62,63], and dam displacement monitoring [64–66] but also detect data anomalies caused by sensor faults or environmental changes [67–69], and even predict missing data [70–73].”
BackgroundA Data Loss Recovery Technique Using EMD-BiGRU Algorithm for Structural Health Monitoring
Liu, D., Bao, Y., He, Y., & Zhang, L. (2021) · Applied Sciences · doi:10.3390/app112110072 · Cited by 23 · cited 1×
“Considering the small dataset and class imbalance problem mentioned above, the possibility of errors in the output of DL algorithms cannot be ignored.”
BackgroundMulti‐resolution broad learning for model updating using incomplete modal data
Kuok, S., & Yuen, K. (2020) · Structural Control and Health Monitoring · doi:10.1002/stc.2571 · Cited by 22 · cited 2×
“The technique used two sets of data from the IASC-ASCE benchmark data, case 1 (undamaged) and case 9 (severely damaged), to train 12 CNN classifiers.”
BackgroundMethodLydon, D., Kromanis, R., Lydon, M., Early, J. M., & Taylor, S. E. (2022) · Journal of Civil Structural Health Monitoring · doi:10.1007/s13349-022-00617-w · Cited by 22
Parziale, M., Lomazzi, L., Giglio, M., & Cadini, F. (2022) · Structural Control and Health Monitoring · doi:10.1002/stc.3076 · Cited by 21 · cited 2×
“…be solved using a combination of the CNN and RNN methods, such as CNN with GRU [64,109,123], CNN with LSTM [27,178–182], CNN with Auto-encoder and [183], CNN with echo state networks [113], etc. Yang et al. [123] proposed a hierarchical CNN and GRU framework called HCG, which used a CNN to…”
MethodGao, X., Huang, C., Teng, S., & Chen, G. (2022) · Applied Sciences · doi:10.3390/app12189244 · Cited by 21
A Deep Learning-Based Intelligent Quality Detection Model for Machine Translation
Chen, M. (2023) · IEEE Access · doi:10.1109/access.2023.3305397 · Cited by 21
Du, B., Lin, C., Sun, L., Zhao, Y., & Li, L. (2022) · IEEE Internet of Things Journal · doi:10.1109/jiot.2022.3141417 · Cited by 20
Sandhu, H. K., Bodda, S. S., & Gupta, A. (2023) · International Journal of Pressure Vessels and Piping · doi:10.1016/j.ijpvp.2022.104849 · Cited by 19
Attard, L., Debono, C. J., Valentino, G., & Castro, M. D. (2021) · Sensors · doi:10.3390/s21124040 · Cited by 18 · cited 2×
“Sajedi et al. [248] proposed that Deep Generative Bayesian Optimization (DGBO) could reduce the number of sensors used for damage localization and evaluation by 52% and 43%, respectively.”
BackgroundMethodDamage analysis and quantification of RC beams assisted by Damage-T Generative Adversarial Network
Qi, Y., Yuan, C., Li, P., & Kong, Q. (2022) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2022.105536 · Cited by 18
Hybrid-Supervised-Learning-Based Automatic Image Segmentation for Water Leakage in Subway Tunnels
Qiu, D., Liang, H., Wang, Z., Tong, Y., & Wan, S. (2022) · Applied Sciences · doi:10.3390/app122211799 · Cited by 18 · cited 1×
“Qiu et al. [174] proposed that WRDeepLabV3+ combined with the class activation map could accurately identify leakage in subway tunnels, which could segment the leakage area more thoroughly.”
BackgroundLi, S., Liu, F., Peng, G., Cheng, F., Zhao, B., & Ji, M. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3214300 · Cited by 18 · cited 3×
“In the studied literature, vibration-based SHM is mainly used in crack [109], brace damage [110–112], bolt loosening [113,114], and stiffness reduction [115,116] detection.”
BackgroundMethodOh, B. K., Jung, W. C., & Park, H. S. (2023) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2023.106019 · Cited by 18
Multi-Point Displacement Monitoring Based on Full Convolutional Neural Network and Smartphone
Zhang, Y., Zhao, X., & Liu, P. (2019) · IEEE Access · doi:10.1109/access.2019.2943599 · Cited by 16 · cited 1×
“The SHM method extracts damage-sensitive features from AE and ultrasonic signals using DL methods such as CNN for DD. Hesser et al. [20] formed three different AE sources using the Hsu–Nielsen pencil core fracture source, 3 mm, and 5 mm ball impact.”
MethodProbabilistic cracking prediction via deep learned electrical tomography
Chen, L., Gallet, A., Huang, S., Liu, D., & Smyl, D. (2021) · Structural Health Monitoring · doi:10.1177/14759217211037236 · Cited by 16 · cited 1×
“Chen et al. [152] used FE to simulate resistance tomography data and an ANN to predict the cracking patterns of reinforced concrete members.”
MethodA Novel Structural Damage Identification Method Using a Hybrid Deep Learning Framework
He, Y., Huang, Z., Liu, D., Zhang, L., & Liu, Y. (2022) · Buildings · doi:10.3390/buildings12122130 · Cited by 16
IterLUNet: Deep Learning Architecture for Pixel-Wise Crack Detection in Levee Systems
Panta, M., Hoque, M. T., Abdelguerfi, M., & Flanagin, M. C. (2023) · IEEE Access · doi:10.1109/access.2023.3241877 · Cited by 15 · cited 1×
“Panta et al. [39] proposed an iterative loop U-Net to segment crack images of levees into crack pixels and non-crack pixels to identify crack shapes accurately.”
BackgroundEffect of Untreated Surface Disbonds on Performance of FRP-Retrofitted Concrete Beams
Kalaycı, A. S., Yalim, B., & Mirmiran, A. (2009) · Journal of Composites for Construction · doi:10.1061/(asce)cc.1943-5614.0000032 · Cited by 13
Morales‐Valdez, J., Lopez-Pacheco, M. A., & Yu, W. (2020) · Structural Control and Health Monitoring · doi:10.1002/stc.2584 · Cited by 13
Pitch-catch UGW-based multiple damage inference: a heterogeneous graph interpretation
Zhou, L., Chen, S., Ni, Y., & Liu, J. (2021) · Smart Materials and Structures · doi:10.1088/1361-665x/ac36b0 · Cited by 13 · cited 1×
“Nguyen et al. [218] developed a 1D-CNN model to extract and learn the optimal features from the original electromechanical impedance signals to assess the severity of prestress loss in reinforced concrete beams.”
MethodLiu, B., Xu, Q., Chen, J., Li, J., & Wang, M. (2022) · Buildings · doi:10.3390/buildings12081286 · Cited by 13 · cited 3×
“…and Gated Recurrent Unit (GRU), it can not only realize multi-type damage identification and location [61], crack detection [62,63], and dam displacement monitoring [64–66] but also detect data anomalies caused by sensor faults or environmental changes [67–69], and even predict missing data [70–73].”
BackgroundDetection of Nut–Bolt Loss in Steel Bridges Using Deep Learning Techniques
Li, Z., Adamu, K., Yan, K., Xu, X.-L., Shao, P., Li, X.-H., & Bashir, H. M. (2022) · Sustainability · doi:10.3390/su141710837 · Cited by 13
Feature-based hybrid strategies for gradient descent optimization in end-to-end speech recognition
DOKUZ, Y., & TÜFEKCİ, Z. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-022-12304-5 · Cited by 12
Energy Ratio Variation-Based Structural Damage Detection Using Convolutional Neural Network
Wu, C., Peng, Y.-X., Zhuo, D.-B., et al. (2022) · Applied Sciences · doi:10.3390/app122010220 · Cited by 12
Crack Location and Degree Detection Method Based on YOLOX Model
Wang, L., Li, J., & Kang, F. (2022) · Applied Sciences · doi:10.3390/app122412572 · Cited by 12 · cited 4×
“…can make use of structural surface defect images of bridges, frame structures, dams, tunnels, and other infrastructures for crack detection [40–44], bolt loosening detection [45–47], rebar surface defect detection [48], delamination and reinforcement exposure detection [49], displacement…”
BackgroundInfluentialOh, B. K., & Kim, J. (2021) · IEEE Access · doi:10.1109/access.2021.3057687 · Cited by 11 · cited 1×
“However, whether the method of bridge crack identification can be applied to dam crack identification or even bridge deformation identification is a concern.”
BackgroundMa, Z., & Gao, L. (2021) · KSCE Journal of Civil Engineering · doi:10.1007/s12205-021-1307-z · Cited by 11
Deep Learning Approach to Impact Classification in Sensorized Panels Using Self-Attention
Karmakov, S., & Aliabadi, M. H. (2022) · Sensors · doi:10.3390/s22124370 · Cited by 11
Pham‐Bao, T., Ngo, N. K., Vuong‐Cong, L., & Nguyen‐Nhat, T. (2022) · Structural Control and Health Monitoring · doi:10.1002/stc.2956 · Cited by 9 · cited 1×
“Then, input acceleration and lag displacement to the CNN classifier can improve the accuracy of the damage location.”
MethodDeep Learning-Enriched Stress Level Identification of Pretensioned Rods via Guided Wave Approaches
Zhang, Z. M., Tang, F., Cao, Q., Pan, H. X., Wang, X., & Lin, Z. (2022) · Buildings · doi:10.3390/buildings12111772 · Cited by 9
Xiao, H., Ogai, H., & Wang, W. (2024) · Structure and Infrastructure Engineering · doi:10.1080/15732479.2023.2167214 · Cited by 9
Application of C-LSTM Networks to Automatic Labeling of Vehicle Dynamic Response Data for Bridges
Shin, R., Okada, Y., & Yamamoto, K. (2022) · Sensors · doi:10.3390/s22093486 · Cited by 8
A high‐frequency feature enhancement network for the surface defect detection of welded rebar
Li, Z., & Li, D. (2022) · Structural Control and Health Monitoring · doi:10.1002/stc.2983 · Cited by 8 · cited 1×
“…defect images of bridges, frame structures, dams, tunnels, and other infrastructures for crack detection [40–44], bolt loosening detection [45–47], rebar surface defect detection [48], delamination and reinforcement exposure detection [49], displacement detection [50–53], voids detection [54], etc.”
BackgroundNeural network-based prediction of topside mass of an in-service jacket platform
Huang, Y. Q., & Huang, S. (2022) · Ocean Engineering · doi:10.1016/j.oceaneng.2022.110554 · Cited by 7 · cited 1×
“Other practical structures include the dam, tunnel, railroad, and Canton Tower, the data of which can verify the effectiveness of the proposed method.”
MethodSakurai, H., Suzuki, K., Ishii, S., et al. (2022) · Nuclear Materials and Energy · doi:10.1016/j.nme.2022.101171 · Cited by 6
Semi-supervised classification of hyperspectral images based on multi-view consistency
Liu, B., Zuo, X., Yu, A., et al. (2023) · Remote Sensing Letters · doi:10.1080/2150704x.2023.2215946 · Cited by 6
Learning Damage Representations with Sequence-to-Sequence Models
Yang, Q., & Shen, D. (2022) · Sensors · doi:10.3390/s22020452 · Cited by 5 · cited 1×
“In the studied literature, vibration-based SHM is mainly used in crack [109], brace damage [110–112], bolt loosening [113,114], and stiffness reduction [115,116] detection.”
MethodResearch on English Vocabulary and Speech Corpus Recognition Based on Deep Learning
Zhen, W. (2022) · Wireless Communications and Mobile Computing · doi:10.1155/2022/2882964 · Cited by 3
Structural Control: Past, Present, and Future
(1997) · doi:10.1061/(asce)0733-9399(1997)123:9(897
Research works printed but not linked: 3
- Rytter, A. (1993). Vibrational Based Inspection of Civil Engineering Structures. [Ph.D. Thesis, Aalborg University].
- Hosoda, 2022, Prediction of maximum crack width by machine learning using concrete construction data in yamaguchi system, Int. J. Intell. Inform. Infrastruct. Data Sci. Pap., 3, 898
- Grieves, M., and Vickers, J. (2017). Transdisciplinary Perspectives on Complex Systems: New Findings and Approaches, Springer.
Cited by
AI in structural health monitoring for infrastructure maintenance and safety
Plevris, V., & Papazafeiropoulos, G. (2024) · Infrastructures · doi:10.3390/infrastructures9120225 · Cited by 132
Ghazimoghadam, S., & Hosseinzadeh, S. A. A. (2024) · Measurement · doi:10.1016/j.measurement.2024.114410 · Cited by 107
Eddy Currents Probe Design for NDT Applications: A Review
Machado, M. A. (2024) · Sensors · doi:10.3390/s24175819 · Cited by 82
“SHM integrates SHM integrates NDT methods with sensors [32], data analysis algorithms [33,34], and structural models to provide real-time information on structural health [35,36], enabling proactive maintenance and minimizing the risk of catastrophic failure [37–41].”
MethodDigital twin systems for musculoskeletal applications: A current concepts review
Diniz, P., Grimm, B., Garcia, F., Fayad, J., Ley, C., Mouton, C., Oeding, J. F., Hirschmann, M. T., Samuelsson, K., & Seil, R. (2025) · Knee Surgery, Sports Traumatology, Arthroscopy · doi:10.1002/ksa.12627 · Cited by 100
Sivasuriyan, A., Vijayan, D. S., Devarajan, P., Stefańska, A., Dixit, S., Podlasek, A., Sitek, W., & Koda, E. (2024) · Sensors · doi:10.3390/s24248161 · Cited by 58
Metadata
- Authors
- Jia, J., & Li, Y.
- Affiliations
- Ocean University of China (CN)
- Corresponding author
- Jia, Jing
- Venue
- Sensors
- Publisher
- MDPI AG
- Published
- 2023-10-30
- Volume
- 23
- Issue
- 21
- Pages
- 8824
- ISSN
- 1424-8220 (ISSN-L)
- PMC
- PMC10650096
- OpenAlex
- W4388018712
- Semantic Scholar
- 2eb5a7ef26f6bd0da7f824db30d3882e7c5a1ace
- Cited by
- 182
- References
- 271
- Citations by year
- 2026: 97 · 2025: 64 · 2024: 20 · 2023: 1
- FWCI
- 3.31
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Sensors (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- National Natural Science Foundation of China (51908523)
- Topics
- Physical Sciences › Engineering › Civil and Structural Engineering › Structural Health Monitoring Techniques
- Physical Sciences › Engineering › Civil and Structural Engineering › Infrastructure Maintenance and Monitoring
- Physical Sciences › Engineering › Civil and Structural Engineering › Structural Engineering and Materials Analysis
- Keywords
- structural health monitoring, deep learning, damage detection, crack detection, vibration signals, image-based damage detection, convolutional neural networks, digital twin, bridge structures, structural damage localization
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-05, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-05 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.