Heilongjiang Provincial Department of Transportation Big Data Platform Construction Project

Designed to fully leverage the data center network infrastructure of Heilongjiang Provincial Department of Transportation, the platform integrates information resources from various business systems using advanced data connectivity technologies, creating a provincial transportation big data service platform that offers centralized data, comprehensive catalogs, clear themes, convenient functionality, and thoughtful services.

As the province’s transportation information resource hub, the platform breaks down departmental silos by integrating data from urban traffic management, the broader transportation industry, and other sectors. It aggregates diverse traffic data from highway administrations, road administrations, enterprises, institutions, and internet sources, enabling the integration, sharing, analysis, computation, and interaction of multi-source heterogeneous data. This supports comprehensive and in-depth mining and utilization of transportation information, providing backend support for high-quality, efficient traffic management and services.

The big data platform for Heilongjiang Provincial Department of Transportation has deployed 4 nodes, with a total data volume exceeding 8TB, daily increment over 50GB, nearly 8,000 data tables, more than 2,000 stored procedures, and 300 concurrent users. The platform is now officially online and has entered a stable phase. It will undertake the task of supporting data access for all secondary departments in the transportation industry across Heilongjiang Province.

The overall technical architecture of the big data platform consists of four parts:

• Data Sources: The data resources of the big data platform include the provincial data resource center and data resource partitions for highway management, road transportation management, waterway management, and comprehensive government affairs.

• Data Integration Layer: Its main role is to collect data from various data sources, and through processes such as data cleansing and comparison, realize data import, aggregation, organization, and querying. In this platform, data integration accomplishes two main tasks: First, complete data integration for the provincial department’s data warehouse. The provincial data warehouse primarily extracts basic data, thematic data, and shareable business data from various data partitions of transportation industry management departments. Second, extract useful data from business systems in domains such as highway management, road transportation management, industry management, and comprehensive government affairs, and after cleansing and transformation, load them into the partitioned data warehouses to support data statistical analysis and management within those domains.

• Data Resource Layer: This includes the data warehouse and supporting repositories such as shared information repository, rule base, model base, metadata repository, and unstructured data metadata repository. The GBase 8a MPP Cluster is primarily used to store all data after integration, including basic databases, business databases, thematic databases, shared databases, and multiple data analysis thematic databases divided according to analytical business requirements, such as thematic databases for highway transportation travel patterns and operation monitoring, highway safety accident monitoring, long-distance passenger transport operation monitoring, etc.

• Application Layer: This includes an application support system and business analytics applications. The application support system provides various engines for upper-layer applications, including ad-hoc query, multidimensional analysis, interactive charts, interactive reports, geographic information engine, permission management, semantic mapping, model design, etc., supporting upper-layer application development. Business analytics applications are built on the visualized data analysis system, constructing analysis applications based on data themes, including unified query, comprehensive analysis, data prediction, and other business analyses.

Figure 1: Business Architecture Diagram of the Big Data Analytics Platform

The Heilongjiang Provincial Department of Transportation built its transportation big data platform with GBase 8a MPP Cluster, achieving dual benefits in both technology and business.

• Significantly reduced construction costs: The entire platform is built on commodity x86 servers. With 1:8 data compression on ingestion, local storage utilization is optimized, significantly reducing data storage costs. The low hardware cost provides ample budget headroom for future expansion. As the platform's data volume increases, GBase 8a MPP's superior cost-effectiveness will stand out.

• High availability: GBase 8a MPP Cluster’s multi-tier high availability technology safeguards all core data for the Heilongjiang transportation big data platform, ensuring 24/7 stable operation with no single point of failure.

• Secure and controllable: Core data is managed by a domestically developed distributed database with full intellectual property rights, offering self-developed, secure, and controllable features, thereby substantially improving data security and business security of existing transportation data.