A wireless and low power implementation of a new approach for structural Health Monitoring

Structural Health Monitoring (SHM) aims at detecting damage in various infrastructures. Measuring the electro-mechanical coupling between the system under test and a piezoelectric transducer applied to the structure could be an effective approach for SHM. This method relies on the estimation of the system eigenfrequencies. Considering the different scenarios where this technique can be applied and the different structures that can be monitored, the implementation of this method on a wireless and battery-powered embedded system is particularly interesting. In this paper a study on the implementation of eigenfrequencies estimation on a low-power wireless microcontroller is presented. The main challenge to be faced with this hardware platform is the low amount of memory that can be dedicated to the task and the poor parallelization capability. To evaluate the system eigenfrequencies different FFT architectures can be exploited. In this paper, a study is presented to assess the accuracy in the eigenfrequencies detection, the spectral resolution, and the memory requirement. A performance comparison of different architectures is presented, allowing to select the best trade-off and evaluate the feasibility of the implementation of this new approach on a wireless sensor node based on a low-power microcontroller.