This system is based on environmental sound recognition technology, using Arduino to develop low-cost smart sensors and deploying the LRF-MobileNet deep learning model. Sensors pre-installed with machine learning models parse equipment status data in realtime and upload it to servers. Managers achieve centralized equipment monitoring and hazard warnings through web/mobile terminals, forming a closed-loop operation and maintenance pattern of "edge perception-cloud management".

Technology provider:Nanjing University of Posts and Telecommunications