Digital Twin Synchronization with Real Time Data Collection

Digital twin systems rely heavily on accurate and time-synchronous data synchronization between physical entities and their virtual counterparts. However, in industrial environments, real-time data acquisition processes present significant challenges due to communication delays, batch data transfer, and temporal mismatches between the sensing and processing layers. This study focuses on the data and synchronization layer of the digital twin architecture and proposes a timestamp-based synchronization framework aimed at ensuring temporal consistency under near real-time conditions. The proposed approach has been validated on an overhead crane system where load and rope angle data are acquired. Sensor measurements are collected at 100 ms intervals on a Raspberry Pi Compute Module-based edge device but transmitted to the server with configurable batch transmission intervals. In the current system, this interval is set to 5 seconds. To address transmission delay and temporal mismatch issues, each sensor measurement is timestamped on the edge device and used for reconstruction of the digital twin state on the server side. The proposed synchronization framework prioritizes temporal consistency over instantaneous updating by utilizing buffering, time alignment, and state reconstruction mechanisms. This ensures that the digital twin accurately reflects the physical system state within a certain delay limit, despite limited communication frequency. Experimental evaluations conducted under varying data transmission intervals demonstrate that the proposed method reduces synchronization error and maintains data consistency. The results show that reliable near-real-time digital twin synchronization is possible in industrial systems where high-frequency sensing and delayed data transmission are used in combination, using a timestamp-based reconstruction approach.