Abstract
<title>Abstract</title> <p>The convergence of Artificial Intelligence (AI), Internet of Things (IoT), and Digital Twin (DT) technologies is revolutionizing modern manufacturing, yet existing implementations often remain fragmented, lacking real-time adaptability, human-centric integration, and scalable architecture. This study proposes a novel five-layer AI–IoT–DT framework designed to enable intelligent, sustainable, and human-aligned manufacturing systems. The architecture incorporates real-time data acquisition through IoT sensors, dynamic digital twin modeling for simulation and control, machine learning-driven analytics for predictive maintenance, AR-enabled interfaces for human–machine collaboration, and closed-loop feedback for continuous optimization. Validated in a Python-based simulation environment using synthetic industrial datasets, the framework demonstrated high predictive accuracy with LSTM models (F1-score: 0.93), a 27% reduction in unscheduled downtime, a 12.4% decrease in energy consumption per unit, and a 17% improvement in Mean Time Between Failures (MTBF). Additionally, operator cognitive load was reduced by 19% through AI-assisted interfaces. Comparative analysis with prior studies highlights the framework’s superior performance and unique emphasis on Industry 5.0 values—such as sustainability, human inclusion, and ethical AI deployment. This research offers a scalable, modular blueprint for transitioning from traditional automation to intelligent augmentation in manufacturing. Future work will focus on real-world deployment, integration with cybersquatting and edge computing, and adherence to international digital twin standards.</p>