@article{DENIS2026131259,
title = {OxiMoRa: Oxidation Monitoring of Electrochemical Steel Corrosion based on Harmonic Radar and Deep Learning},
journal = {Expert Systems with Applications},
pages = {131259},
year = {2026},
issn = {0957-4174},
doi = {https://doi.org/10.1016/j.eswa.2026.131259},
url = {https://www.sciencedirect.com/science/article/pii/S0957417426001727},
author = {Nathanael Denis and Roberto {Di Pietro}},
keywords = {Harmonic Radar, Oxidation, Steel Corrosion, Corrosion Monitoring, Machine Learning, Structural Health Monitoring},
abstract = {Harmonic radar leverages the non-linear electromagnetic response of materials to detect and monitor specific phenomena, offering a unique perspective on corrosion monitoring in metals. Although harmonic radar has been explored for detecting electronics and batteries, its application to steel corrosion monitoring has been limited, despite the semiconducting behavior of rust oxides. To address this gap, we propose OxiMoRa, a novel methodology for wireless short-range steel oxidation monitoring that combines harmonic radar with machine learning. Our methodology is validated on steel corrosion scenarios using a low-cost custom-built harmonic radar testbed and software-defined radios, demonstrating its capability to detect corrosion. Extensive experimental results show robust performance, achieving a minimum F1-score of 0.96 in line-of-sight detection and of 0.86 in challenging classification scenarios. These excellent results pave the way for further research on the usage of harmonic radar for corrosion monitoring and structural integrity checks.}
}