A Low-Complexity Envelope-Based Method for Efficient Sensing and Communication in Remote Vibration-Based Condition Monitoring

In remote sensing and condition monitoring applications, integrated sensing and communications (ISAC) can enhance process efficiency by providing real-time communication and sensing for operational and environmental conditions. Further, recent advances in sensor technology facilitate the use of smart transducers, where the acquired signal is analyzed and processed at the measurement point, thereby improving overall efficiency. However, the limited processing capabilities of smart transducers, in general, necessitate the development of efficient algorithms to process the signal and extract relevant data at the measurement point. To this end, we introduce a low-complexity method for vibration-based condition monitoring applications, utilizing envelope analysis of acquired vibration signals. In the proposed method, the instantaneous amplitude (envelope) and instantaneous frequency of the acquired vibration signal are jointly utilized to facilitate a new spectral representation of the signal’s envelope. Maintaining instantaneous information, the introduced representation effectively captures energy-frequency variations that are unique to the current operational condition, thereby enabling the extraction of a few discriminative features that are sensitive to changes in operational condition. The developed method is evaluated using a rolling bearings vibrations dataset, and the experimental results demonstrate excellent performance in detecting and diagnosing various fault types. Additionally, the proposed method has moderate computational complexity, meeting the computational requirements of real-time monitoring and smart transducers.