Performance Evaluation and Field Validation of Next-Generation QMEMS Accelerometers for Seismology, Structural Health...

Recent advances in Micro-Electro-Mechanical Systems (MEMSs) have enabled the development of accelerometers increasingly suitable for seismological and structural engineering applications. Quartz MEMS (QMEMS) sensors combine low self-noise, wide dynamic range, excellent thermal stability, and compact dimensions, providing a cost-effective alternative to conventional force-balance and piezoelectric accelerometers. This study presents the development and validation of a complete QMEMS-based sensing platform for seismic monitoring and Structural Health Monitoring (SHM), integrating the recently introduced Epson M-A370 accelerometer, a Smart Sensor Box with precise timing synchronization, and embedded acquisition and edge-processing capabilities. The platform was evaluated through comprehensive laboratory and field experiments. The M-A370 was experimentally compared with the M-A352 and a reference force-balance accelerometer, while complementary M-A352 tests included representative MEMS and piezoelectric accelerometers. The experimental results indicate that accelerometer self-noise is a primary factor governing the reliability of Operational Modal Analysis (OMA) and long-term SHM. Self-noise densities below 1 μg/√Hz, preferably below 0.5 μg/√Hz, and, for the most demanding applications, approaching or below 0.1 μg/√Hz, represent practical performance targets for robust modal identification and reliable tracking of structural dynamic properties under weak ambient excitation or in very quiet environments. These values, however, are not exclusion thresholds, as higher-noise accelerometers can reliably record ground motions sufficiently above their instrumental noise floor. The ultra-low-noise M-A370 (0.02 μg/√Hz) delivers data quality comparable to engineering-grade force-balance accelerometers. The proposed platform, combining ultra-low-noise QMEMS technology, precise timing synchronization, and embedded processing, provides a scalable framework for Urban Seismic Observatories, distributed SHM, OMA, and impact-based Earthquake Early Warning (EEW) across buildings, bridges, and heritage structures.