Researchers have developed a lightweight AI framework for monitoring the condition of consumer devices using vibration data. The system converts 1D vibration signals into 2D time-frequency representations using CWT and HHT, which are then analyzed by lightweight deep learning models like MobileNetV3-Small, ConvNeXt-Tiny, and YOLOv11-cls. Tested on public bearing datasets from Case Western Reserve University, Paderborn University, and Intelligent Maintenance Systems, the framework aims for efficient embedded deployment in appliances like washing machines and robotic vacuums. Results indicate that YOLOv11-cls offers high diagnostic accuracy, while MobileNetV3-Small provides a low-latency option for edge platforms, highlighting trade-offs between accuracy, computational cost, and inference speed. AI
IMPACT Enables more efficient and accurate predictive maintenance in resource-constrained consumer electronics.
RANK_REASON Academic paper detailing a novel AI framework for signal processing and condition monitoring. [lever_c_demoted from research: ic=1 ai=1.0]
- alphaXiv
- Case Western Reserve University
- CatalyzeX Code Finder for Papers
- ConvNeXt-Tiny
- DagsHub
- Gotit.pub
- Hugging Face
- Intelligent Maintenance Systems
- International Musicological Society
- MobileNetV3-Small
- Purdue University
- ScienceCast
- University of Paderborn
- YOLOv11-cls
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
