With the growing concern of aging infrastructures, the need for effective and non-intrusive monitoring techniques has become increasingly important. While most current methods rely on active testing, this study explores the potential of using ambient noise interferometry as a passive method for Structural Health Monitoring (SHM) of concrete structures. It investigates whether the Green's function (GF) of concrete medium can be adequately estimated from traffic noise to assess its health condition.

Two datasets are examined: a validation dataset from a laboratory experiment simulating ambient noise on a pre-stressed concrete girder, and real-world traffic noise data from the Maastunnel in Rotterdam. For each dataset, the following aspects are analyzed: (1) signal characteristics, including amplitudes and frequency distributions; (2) the optimal pre-processing scheme, incorporating temporal and spectral normalization, along with frequency filtering; and (3) the coherence of the resulting GF estimation from interferometry, particularly time of wave arrivals.

The results from the validation dataset demonstrate that ambient noise interferometry can reliably reconstruct the GF for concrete medium, indicating its effectiveness for monitoring changes such as crack formation and strain changes. However, the analysis of actual traffic noise data did not provide sufficient evidence to support its use for SHM with the current setup. Although a coherent and usable frequency range for traffic noise was identified, the limited amount of data led to a low signal-to-noise ratio (SNR), which made it challenging to highlight relevant features.

Moving forward, future researchers are encouraged to collect sufficient amount of data for analysis to better determine the feasibility of reconstructing the GF with ambient traffic noise. Additionally, exploring alternative sampling methods like continuous recording could address one of the limitations of this research. Finally, employing decomposition methods may help in increasing the SNR.