This article examines the control, diagnostics, and optimization of mining electromechanical systems based on artificial intelligence technologies. Modern mining enterprises require high reliability of electric drives, pumping units, conveyor systems, and hoisting mechanisms. Traditional maintenance methods are reactive in nature and often lead to unexpected failures and downtime. The study analyzes the application of machine learning algorithms for predictive diagnostics using vibration monitoring, thermal control, and electrical parameter analysis. Artificial neural networks and regression models were applied to process operational data and identify hidden patterns affecting system performance. The research results demonstrate that the implementation of artificial intelligence-based predictive maintenance systems increases fault detection accuracy up to 92–96%, reduces energy consumption by 12–18%, and decreases emergency situations by 30–40%. The findings confirm the effectiveness and перспективeness of intelligent control systems in improving safety, reliability, and energy efficiency in mining enterprises.