Towards Efficient Structural Health Monitoring: Spike-Encoding in Real-Time Electromechanical Impedance Systems

In this work, we introduce the Spike Count Deviation (SCD) algorithm for use with spike-encoded electromechanical impedance (EMI) data, aimed at enhancing on-edge structural health monitoring (SHM) systems. The proposed approach achieves significant data compression without compromising accuracy. We apply step-forward spike encoding alongside the SCD algorithm to two reinforced concrete samples under five simulated damage scenarios. Experiments show a damage detection accuracy of 99.53%, with an average of 350 generated binary spikes.The performance of the proposed approach is compared to two studies using the conventional Root Mean Square Deviation (RMSD) algorithm for damage detection. We demonstrate a comparable detection accuracy of >99% and show that the SCD algorithm is resistant to temperature effects, exhibiting 10 times greater sensitivity to damage than to temperature variations.