<![CDATA[Bridge structures undergo continuous degradation due to traffic loading and environmental exposure, necessitating the development of practical methods to monitor changes in their dynamic response. This study examines the use of a low-cost MEMS accelerometer for identifying dominant modal frequency bands (natural-frequency candidates) of an operational short-span bridge under ambient excitation. An ADXL345 sensor, integrated with an Arduino-based data acquisition system and a MATLAB interface, was used to record tri-axial vibration signals at three locations on the Jembatan Jalan Gelatik in Samarinda during both daytime and late-afternoon traffic conditions. The time-domain signals were processed using Welch’s averaged windowed Fast Fourier Transform, followed by Konno–Ohmachi smoothing to clarify local spectral peaks. The analysis was intentionally limited to frequencies below 20 Hz, where global modes are expected, and the signal-to-noise ratio of the MEMS sensor is more reliable. Several consistent modal frequency bands were identified across measurement points, with dominant peaks observed between approximately 1.3–1.5 Hz, 2.1–2.7 Hz, 3.3–3.5 Hz, 5.0–6.8 Hz, 8.0–9.0 Hz, and 14–18 Hz. These peaks were validated through spatial repeatability across measurement points and temporal repeatability across different traffic conditions (daytime and late afternoon). These results indicate that the combination of low-cost sensing and noise-robust spectral processing can extract stable modal information from ambient bridge vibrations, despite the limitations of single-sensor deployment and the absence of reference-grade instruments. The findings suggest that this approach offers a feasible preliminary method for vibration-based structural assessment and may serve as a foundation for further development toward more detailed modal characterization.]]>
