International Scientific Journals
of Scientific Technical Union of Mechanical Engineering "Industry 4.0"

The aim of this work is to implement a wireless communication system for MEMS-based sensors within the framework of Internet of Things (IoT) applications, specifically in the context of Predictive Maintenance (PdM). The focus is placed on developing a functional prototype of a wireless sensor node that enables efficient data acquisition and transmission from commercially available MEMS vibration sensors. The solution leverages an ESP32 microcontroller for data handling and Wi-Fi communication, forming the basis of a scalable wireless sensor network (WSN). The project emphasizes their integration into a wireless system architecture suitable for industrial monitoring scenarios. This approach aims to demonstrate how low-cost MEMS sensors, when combined with IoT technologies, can contribute to accessible and modular condition monitoring solutions aligned with Industry 4.0.

This study evaluates the performance of newly designed deep-learning model—bidirectional long short-term memory network (Bi-LSTM), with baseline to a conventional multilayer perceptron (MLP)—for classification faults of rolling-element bearings from raw vibration signals. The models are benchmarked against a previously optimised one-dimensional convolutional neural network (1-D CNN), originally obtained via Bayesian hyperparameter search. A carefully selected dataset of 3600 one-second segments was captured under varying speed conditions and dynamically enhanced with Gaussian noise during processing. On the test set, the Bi-LSTM achieves 100 % accuracy, the 1-D CNN 97.9 %, and the MLP 53.3 %. Training dynamics, confusion patterns, and model complexity were thoroughly analysed, highlighting the trade-offs between accuracy, latency and deployment cost in edge-computing scenarios.