Metrological evaluation of new industrial SHM systems based on MEMS and microcontrollers

Microcontrollers are becoming popular in Structural Health Monitoring (SHM) systems, as they can manage sensors, process data and meet requirements complying to the cloud and Internet of Things (IoT) paradigms. Similarly, micro electromechanical system (MEMS) sensors are spreading for monitoring applications, given their appealing costs. Considering the importance of data for a SHM-targeted decision-making process, a conscious use of these technologies requires a deeper analysis from a metrological point of view, to ensure reliability and robustness of the provided data. Consequently, in a multi-node sensing architecture, concepts like sensitivity of each node or data synchronization becomes of uttermost importance, especially if modal parameters extraction is sought. This paper is intended as a warning for designers of SHM monitoring architectures: can we simply replace standard sensing devices with low-cost systems and expect they perform as well as their “stronger brothers”? The answer to this question is tentatively provided by discussing results obtained in a testing campaign performed on some reinforced concrete beams, dynamically tested through two different monitoring systems: a standard, high-performance system exploiting high-sensitivity piezoelectric accelerometers and a low-cost MEMS digital accelerometers-based one, not coupled to a high performance data acqusition system, rather. Modal parameters are considered as the target measure to assess the performance of the two systems.