Embedding SHM Algorithms into a Microcontroller for Real-Time and Fully-Automated Civil Applications

The feasibility of long-term, real-time structural diagnosis is investigated by means of a de-centralized architecture of sensor placement. The ultimate goal is that each sensor device communicates the data on the structural status, only when the response re-elaboration shows anomalous features. The first step of the feasibility study consists of embedding the structural health monitoring (SHM) procedure in a microcontroller, which can be coupled with one or more sensors to form a smart sensor device. The simplifications in the SHM algorithms required by the specific features of the microprocessors, which are compatible with the sensor electronics, are emphasized. A comparison of the performances shown by the numeric and the embedded models is done. In particular, the margins of improvement are outlined.