A Large-Scale Time Series Data Management Platform for Cultural Heritage Site Risk Monitoring: Architecture Design and Application

Storage and analysis of large-scale risk-monitoring time series data is essential for the preventive conservation of cultural heritage sites. However, as these monitoring data accumulate to large scales, traditional relational databases can face serious performance bottlenecks, such as high latency when writing data from multiple sensors and inefficient data retrieval. These problems can significantly hinder the analysis of the risk-monitoring data. To address these issues, we propose a software architecture for a data management platform geared towards large-scale risk-monitoring data for cultural heritage sites. The architecture includes a data bus layer, computing engine layer, data management and storage layer, data analysis and visualization layer, and a cluster management and high-availability services layer. The function of each layer is described, as well as the strategy for data persistence and key business pipelines. Next as a practical test, we use open source software to build a pilot risk-monitoring data management system for the Mogao Grottoes, a world heritage site in Gansu, China. We select the system’s parameters based on performance comparisons and then demonstrate the implementation of its functions such as sensor data integration, risk alerts, and business data interpretation. The results show that the architecture can handle high-volume concurrent data insertion, large-scale data management, and long-term historical data operations. This proposed architecture can serve as a reference for those constructing risk-monitoring data management platforms for other cultural heritage sites and museums.