This chapter explores the potential of Generative AI in enhancing predictive maintenance and fault diagnosis within the context of Industry 4.0. As manufacturing companies strive to improve productivity and reduce downtime, this innovative approach goes beyond traditional predictive maintenance models by utilizing historical failure data, machine learning-based control limits, and optimal sensor thresholds to predict and mitigate issues. The chapter also integrates Standard Operating Procedures (SOPs) and historical maintenance records into a comprehensive diagnostic system. By implementing a Retrieval-Augmented Generation (RAG) system combined with Large Language Models (LLMs), this chapter demonstrates how this approach analyzes sensor data, SOPs, and maintenance logs to generate detailed, context-aware maintenance responses leading to more effective and timely decision-making. The study is illustrated through a simulated pump-related scenario, showcasing the successful application of the proposed methods. The findings reveal significant improvements in identifying and diagnosing equipment anomalies, offering a proactive maintenance strategy that enhances operational reliability and efficiency. By incorporating AI-driven techniques like Skope-Rules and RAG, this chapter highlights the critical role of AI in modernizing manufacturing processes and sets the stage for future research focused on real-time processing and broader equipment monitoring.