AI-Based Predictive Maintenance

System for Industrial Machines

(HVAC-R)

For industries like manufacturing, cold storage, and pharmaceuticals to maintain steady cooling, industrial HVAC refrigeration systems are essential. But because they are always in use, they are vulnerable to malfunctions that result in expensive downtime and inefficient use of energy. Conventional maintenance techniques are frequently inefficient or reactive. A low-cost, AI-based predictive maintenance system for small-to-medium-sized businesses is presented in this paper. The system tracks important parameters like temperature, vibration, and power consumption using real-time data from Internet of Things sensors that are connected to an ESP32 module. High accuracy fault prediction and anomaly detection are achieved by a hybrid machine learning model that combines an LSTM Autoencoder with a Random Forest classifier and regression. A chatbot and mobile app are part of the setup for easy-to-use monitoring. The solution can support more dependable and sustainable HVAC operations by lowering maintenance costs, improving energy efficiency, and reducing unplanned breakdowns.