Implementation of Information Entropy in an Industrial Internet of Things Approach for Structural Health Monitoring Applications

Structural Health Monitoring (SHM) involves damage assessment processes that contribute towards overall safety decisions. The need for real-time assessment and decision-making in SHM has long been attempted in various ways via connections between data acquisition and information extraction. In this context, this investigation presents a novel approach to enable real time data streams for SHM. To achieve this goal, an Industrial Internet of Things (IIoT) framework developed is used in conjunction with Nondestructive Evaluation (NDE) datasets for near real-time diagnostics. To demonstrate the performance and results of applying this method, the case of laboratory scale testing of crack initiation is presented in this manuscript. Specifically, compact-tension specimens of an aerospace-grade aluminum alloy were used in accordance with ASTM standards. Acoustic Emission (AE) datasets were acquired and were subsequently used in an in-house built, scalable IIoT system capable of edge, fog, and cloud computing. At the fog layer, a trained model was loaded to classify the signals in real-time. The trained model relies on signal Information Entropy (IE) values as input and outputs to form an indicator of crack initiation. The AE data input is shown as a test-case for any general time-series type data acquired in SHM applications such as accelerometers and vibration sensors. The main innovation of this approach is the fact that a combination of hardware, computing and IE analysis proves to be advantageous to flag the incubation and subsequent initiation of fracture. The IIoT system described can be applied to a variety of SHM applications for continuous type monitoring.