LOW-COST 5G WIRELESS ACCELEROMETER FOR CONDITION MONITORING

Abstract

Machine and equipment downtime have significant impacts on manufacturing costs, which has resulted in a long history of research into condition-based monitoring (CBM). While high-performance CBM systems exist in academic research settings, industrial implementation is usually limited to highly critical assets, due to high initial investment costs and network limitations. Even when used, industrial systems often employ manual periodic measurements performed by hand, resulting in data being missed that could improve decision making. Wireless sensors, utilizing lower cost modern micro-electrical mechanical system (MEMS) accelerometers, present an opportunity for wider CBM deployment in industry by lowering the required investment. This thesis addresses these barriers by developing and evaluating a low-cost, wireless CBM system compatible with both 5G mmWave and Wi-Fi networks. A sensor prototype was developed using cost-effective MEMS accelerometers, specifically the ADXL-357 and ICM-42688-P, which have not been previously evaluated for CBM applications. The system's performance was compared to a high-end Integrated-electric piezoelectric (IEPE) system using a vibration shaker and a linear motion testbed. Results show that the low-cost MEMS sensors can provide data comparable to the IEPE reference, particularly for low frequency monitoring tasks. The 5G mmWave network performance testing showed that it can support high-throughput, low-latency data streams, with speeds and latencies better than current wireless standards. Overall, this research shows that by combining low-cost MEMS sensors with next-generation wireless networks, it is feasible to create low-cost and scalable real-time wireless CBM systems, bridging the gap between academic research and industrial implementation.