Integrated Digital Twin and Predictive Maintenance for Intelligent Automated Parking

To address fragmented visualization and delayed fault response in traditional Automated Parking Systems, this paper proposes an integrated framework merging Digital Twin (DT) technology with AI-driven Predictive Maintenance. By leveraging a Siemens S7-1500 PLC and the NX environment, the system constructs a high-fidelity 3D model to achieve real-time real-time closed-loop interaction between physical and virtual entities. Large Language Models (LLMs) are introduced to process heterogeneous sensor data within the Industrial Internet of Things (IIoT), enabling precise health assessments and fault predictions. Furthermore, a DAG-based workflow engine automates the entire O&M (Operations and Maintenance) process. Experimental results demonstrate that the proposed framework exhibits superior performance in terms of diagnostic sensitivity and robustness traditional threshold-based methods in both detection accuracy and operational efficiency.