A Synchronization Framework to Minimize Temporal Misalignment in Sensor Networks for Digital Twins

Accurate and precisely synchronized data acquisition is the core challenge for digital twins. Conventional methodologies are often plagued with synchronization issues, leading to temporal misalignment across various sensor nodes. This paper proposes a robust synchronization framework for Internet of Things (IoT) based sensor networks with the aim to achieve simultaneous data acquisition to improve the temporal coherence of the data for digital twins. The proposed framework uses a publish-subscribe model in which sensor nodes are configured as subscribers to a shared topic. A synchronization device periodically publishes a sequence number to the shared topic, prompting the sensor nodes to publish the data tagged with the received sequence number to a common topic. This would ensure seamless data collection to address the temporal misalignment problem in distributed sensor systems. A server-side script processes the sensory data received from different nodes based on the sequence number and appends a time stamp to create a unified dataset. This dataset can be used to train machine learning models for energy consumption prediction and to support a digital twin system for real-time environment visualization. A synchronized dataset not only enhances predictive analytics for energy management but also provides improvements to the realism and responsiveness of digital twin environments. This would support more integrated and accurate digital twins in smart building management and other use cases in which real-time data monitoring is critical.