Lightweight AI improves vibration sensing for smart home devices
Lightweight Time-Frequency Vibration Sensing for Consumer Device Condition Monitoring
Machine Learning
Summary
Many common home appliances like washing machines and fans can be monitored by sensing their vibrations to detect problems early. The authors show how to convert vibration signals into visual patterns that a small and efficient AI model can analyze quickly, even on devices with limited computing power. Their approach works well on real-world datasets and helps balance accuracy with speed and resource use. This means it could be used inside everyday appliances to keep them running smoothly without costly maintenance.
What this means in practice
- •For consumer electronics engineers: Integrate lightweight AI vibration analysis into smart appliances to detect mechanical faults efficiently on limited hardware.$Commercial implications: Enables in-device predictive maintenance features for appliances such as washing machines and fans, improving reliability and reducing service costs.
- •For industrial maintenance teams: Use efficient AI models for fast vibration fault detection on rotating equipment, allowing reduced downtime in resource-constrained environments.
Authors
Po-Heng Chou, Jen-Yu Chiu, Chun-Yu Yeh, Wei-Lung Mao, Walid Saad
Abstract
Consumer rotating devices, including washing machines, air conditioners, electric fans, robotic vacuum cleaners, and home fitness equipment, increasingly rely on vibration sensing for intelligent condition monitoring and predictive maintenance. Ensuring reliable operation in these resource-constrained appliances requires efficient, lightweight artificial intelligence (AI) solutions capable of processing non-stationary signals under real-world operational interference. This paper proposes a lightweight AI-assisted vibration sensing framework that employs continuous wavelet transform (CWT) and Hilbert-Huang transform (HHT) to convert one-dimensional (1-D) vibration time-series signals into two-dimensional (2-D) time-frequency representations. These representations are subsequently analyzed using lightweight deep learning architectures, specifically MobileNetV3-Small, ConvNeXt-Tiny, and the You Only Look Once version 11 classification model (YOLOv11-cls). Public bearing benchmark datasets, including Case Western Reserve University (CWRU), Paderborn University (PU), and Intelligent Maintenance Systems (IMS), are utilized as representative rotating components to investigate model generalization. The proposed framework is designed with eventual embedded deployment in consumer electronics in mind. Experimental results demonstrate that CWT combined with YOLOv11-cls achieves high diagnostic accuracy, while MobileNetV3-Small provides an attractive low-latency alternative for computationally constrained edge platforms. Conventional random splitting is also compared with dataset-specific structured isolation. The results highlight the importance of avoiding data leakage and quantify tradeoffs in classification accuracy, parameter count, floating-point operations (FLOPs), model footprint, preprocessing cost, and central processing unit (CPU) inference latency.