This study proposes a feed-forward neural network (FNN)-based remote fault detection system for mechanical equipment, aiming to identify potential faults early by monitoring and analyzing equipment operation data in real time. The system collects data through sensors, and after preprocessing and feature extraction, the FNN model is used for training and fault diagnosis. Experimental results show that the system can effectively improve the accuracy and efficiency of fault detection and support intelligent maintenance of industrial equipment. This study proposes a feed-forward neural network (FNN)-based remote fault detection system for mechanical equipment, aiming to identify potential faults early by monitoring and analyzing equipment operation data in real time. The system collects data through sensors and uses FNN models for training and fault diagnosis after preprocessing and feature extraction. The experimental results show that: The average response time based on the algorithm experiments is less than 120 ms, and the false alarm and missed alarm rates remain low at 0.4% and 0.2%, respectively, which indicates that the system has a high accuracy in distinguishing between normal and faulty states.