Deep machine learning for detection of acoustic wave reflections
Problem
Acoustic emission damage-detection algorithms can be deceived by reflected waves modulated by structural boundaries, so these reflections must be identified before damage can be located reliably.
Why it matters
Misleading signals could cause incorrect damage estimates and unsafe or costly structural maintenance decisions.
What the AI did
Two deep learning models were trained to classify reflected acoustic emission waves: an LSTM learning directly from time series and a CNN classifying spectrograms.
Results
80% classification accuracy with LSTM and near-perfect accuracy with CNN on Hsu-Nielsen source acoustic emission data.
Impact
Demonstrated only on laboratory-generated acoustic emission data; real-structure deployment not stated.
Record
| Venue | Structural Health Monitoring |
|---|---|
| Published | 2019 |
| Work type | article |
| Authors | Mulugeta Haile, Edward L. Zhu, Christopher D. Hsu, Natasha Bradley |
| Citations | 29 (OpenAlex, at import time) |
| References | 28 |
| Model or method | LSTM recurrent neural network; convolutional neural network on spectrograms |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1177/1475921719881642 |
| Open access | closed |
| Keywords | acoustic emission, deep learning, LSTM, CNN, spectrogram, damage detection |
Importance rating
real but narrower advance
Reflected-wave misidentification can cause incorrect damage estimates and unsafe/costly maintenance decisions; useful for structural health monitoring but demonstrated only on laboratory Hsu-Nielsen data, narrow scope.
More in Engineering & Materials
| # | Paper | Year | Citations | Importance |
|---|---|---|---|---|
| 1 | Scaling deep learning for materials discovery | 2023 | 1,471 |
8
|
| 2 | Application of Machine Learning in Material Synthesis and Property Prediction | 2023 | 165 |
7
|
| 3 | Machine learning-guided discovery of ionic polymer electrolytes for lithium metal batteries | 2023 | 115 |
7
|
| 4 | Machine Learning for Structural Health Monitoring of Aerospace Structures: A Review | 2025 | 44 |
7
|
| 5 | Power System Transient Stability Assessment Based on Machine Learning Algorithms and Grid Topology | 2023 | 42 |
7
|