Predictive maintenance (PdM) leverages data and analytics to anticipate potential failures of system components,enabling preemptive maintenance measures to prevent damage.This approach aims to address predictive challenges in mining equipment maintenance,enhancing equipment reliability and production efficiency.The research process encompasses stages such as data collection,preprocessing,model training,and prediction,as well as decision support and execution.Challenges in implementing PdM with artificial intelligence are analyzed from 3 perspectives:data sources,model transparency and interpretability,and system integration.The findings indicate that AI-based PdM significantly reduces equipment downtime,improves maintenance efficiency,and lowers operating costs.Additionally,the study outlines the application prospects of technologies such as machine learning,IoT,cloud computing,and digital twins in PdM,offering directions for future research.