International Journal of Research and Development in Engineering Sciences https://www.ijrdes.com K. Sujatha 2 6 (November - December) 2020 12 November-December enArtificial IntelligenceDeep LearningIndustrial AutomationHybrid ArchitectureSmart Manufacturing The rapid advancement of Artificial Intelligence AI and Deep Learning DL technologies has transformed industrial automation by enabling intelligent adaptive and selfoptimizing systems This paper proposes a twolevel hybrid architecture for smart industrial automation based on AIdriven deep learning models The architecture integrates both machinelevel and systemlevel intelligence to enhance operational efficiency fault detection predictive maintenance and decisionmaking in real time At the first level deep learning algorithms such as Convolutional Neural Networks CNN and Recurrent Neural Networks RNN are employed for data sensing pattern recognition and anomaly detection from industrial Internet of Things IIoT devices The second level utilizes AIbased control mechanisms including reinforcement learning and expert systems to optimize process automation and resource management The hybrid design ensures scalability adaptability and interoperability across various industrial domains Experimental and analytical evaluations demonstrate improved accuracy reduced downtime and energyefficient performance This approach signifies a step toward achieving fully autonomous intelligent and sustainable industrial ecosystems Copyright (c) 2026 International Journal of Research and Development in Engineering Sciences. All rights reserved. 2020 All Rights Reserved