This research addresses the challenge of optimizing the powder factor (PF) in mining operations to enhance blasting efficiency and sustainability. Machine learning models, including Linear Regression (LR), Random Sample Consensus (RANSAC), and Huber Regressor (HR), were employed to develop a predictive framework for PF optimization. Model performance was evaluated using metrics such as R2 (0.9829 for LR, 0.9827 for RANSAC, and 0.9824 for HR), Mean Absolute Error (MAE), and Root Mean Square Error (RMSE), demonstrating their high predictive accuracy. The findings highlight the capability of machine learning to improve blasting control, reduce environmental impacts, and promote sustainable mining practices. By integrating artificial intelligence into mining operations, this study advances the concept of intelligent mining, offering innovative solutions that enhance operational efficiency while fostering environmental stewardship.