An Embedded Multi-Sensor Architecture for Applications in Structural Health Monitoring

Structural Health Monitoring is mandatory to properly perform health assessment and preventive maintenance planning. Most of approaches available in the literature are based on discrete monitoring, which does not allow for a continuous assessment of the structure status as well as to handle fast alerting in case potential anomalous behaviors are highlighted by instrumental observations. This paper deals with the development of a low-cost multisensor node aimed to provide information on both inertial vibrations and tilt experienced by the investigated structure. Moreover, a dedicated signal processing approach is proposed with the aim of both performing the sensing node assessment and analyzing the frequency content and correlation of recorded signals. Results of different experimental surveys are reported. The estimated resolution of acceleration and tilt measurements are 0.3 mg and 0.016°, respectively.