Smart Manufacturing Systems: AI-Driven Optimization of Industrial Processes

Abstract

Smart manufacturing systems in Industry 4.0 rely on interconnected, data-driven processes to enhance productivity and efficiency. However, conventional rule-based and single-model approaches fail to capture dynamic, high-dimensional industrial data, limiting optimization and adaptability. This research proposes an Artificial Intelligence (AI)-driven model for optimizing industrial processes through intelligent and adaptive decision-making. Multi-source datasets comprising Internet of Things (IoT) sensor streams, machine operation logs, and production performance metrics are utilized. The dataset is preprocessed by using Z-score normalization, and Fast Fourier Transform (FFT) is used for feature extraction to convert time-domain sensor and machine operation data into the frequency domain. The proposed framework utilizes a Non-Dominated Genetic-based Adaptive Light Gradient (NG-ALG), designed to balance exploration and exploitation in industrial optimization while handling conflicting objectives such as cost, energy, and throughput. It integrates Non-dominated Sorting Genetic Algorithm II (NSGA-II) to optimize industrial process parameters by balancing conflicting goals, such as cost reduction, energy efficiency, and production throughput, thereby facilitating intelligent decision-making in Industry 4.0 environments. It also incorporates Adaptive Light Gradient Boosting Machine (Adaptive LightGBM) to provide highly accurate predictions of industrial process behavior and to support AI-driven optimization in smart manufacturing systems. Experimental results demonstrate improved production efficiency, reduced downtime, and enhanced resource utilization with a mean absolute error (MAE) of 0.060, a Mean Squared Error (MSE) of 0.008 and R² of 0.895 which was stimulated in Python. The proposed method enables scalable, adaptive, and data-driven optimization, contributing to intelligent autonomous and efficient smart manufacturing systems.