Abstract: To improve the maintenance efficiency and intelligence level of mechanical and electrical equipment, this paper takes three 7.5 kW CNC lathes in a machining workshop as the research object, and designs and implements an intelligent maintenance system based on LoRa and edge computing. By constructing a multi-source sensor network, deploying an edge computing gateway, and establishing a cloud-based collaborative analysis platform, the system realizes three core functions: Equipment anomaly diagnosis, maintenance cycle prediction, and energy consumption optimization management. Experimental results show that the system achieves the desired performance in terms of anomaly detection accuracy, fault warning lead time, maintenance cycle prediction error, and power factor improvement. The study indicates that the proposed system can effectively enhance equipment operational reliability and reduce maintenance costs, demonstrating promising application potential.