Some construction and engineering companies are already using the concept of Common Data Environment (CDE) according to ISO 19650. In essence, the CDE performs the same functions as a data warehouse (DWH) in other industries: centralizes information, provides version control, and provides access to validated information.
A Common Data Environment (CDE) is a centralized digital space used to manage, store, share and collaborate on project information throughout all phases of the facility lifecycle. CDE is often implemented using cloud-based technologies and integrated with CAD (BIM) systems.
The financial, retail, logistics and industrial sectors have been using centralized data management systems for decades, integrating information from different sources, controlling its relevance and providing analytics. CDE takes these principles further by adapting them to the challenges of building design and lifecycle management.
Like DWH, CDE structures data, captures changes and provides a single point of access to validated information. With the move to the cloud and integration with analytical tools, the differences between the two are becoming less and less apparent. Adding to CDE granular data, the concept of which has been discussed by CAD -vendors since 2023[93, 125], one can see even more parallels with classic DWH.
Earlier in the chapter “Construction ERP and PMIS systems” we have already discussed PMIS (Project Management Information System) and ERP (Enterprise Resource Planning). In construction projects, CDE and PMIS work together: CDE serves as a repository for data including drawings, models and project documentation, while PMIS manages processes such as controlling deadlines, tasks, resources and budgets.
ERP, responsible for business management as a whole (finance, procurement, personnel, production), can integrate with PMIS, providing control of costs and budgets at the company level. For analytics and reporting, DWH can be used to collect, structure and aggregate data from CDE, PMIS and ERP to evaluate financial KPIs (ROI) and identify patterns. In turn, Data Lake (DL) can complement DWH by storing raw and unstructured data (e.g., logs, sensor data, images). This data can be processed and loaded into DWH for further analysis.
Thus, CDE and PMIS focus on project management, ERP focuses on business processes, and DWH and Data Lake focuses on analytics and data operations.
In comparing CDE, PMIS and ERP systems with DWH and Data Lake, significant differences can be seen in terms of vendor independence, cost, integration flexibility, data independence, speed of adaptation to change, and analytical capabilities (Fig. 8.1-11). Traditional systems such as CDE, PMIS, and ERP are often tied to specific vendor solutions and standards, making them less flexible and increasing their cost due to licenses and support. In addition, data in these systems are often encapsulated in proprietary, closed formats, which limits their use and analysis.
In contrast, DWH and Data Lake provide greater flexibility in integrating with different data sources, and their use of open technologies and platforms helps reduce total cost of ownership. Moreover, DWH and Data Lake support a wide range of analytical tools, which enhances analytics and management capabilities.
With the development of reverse-engineering tools for CAD formats and the availability of access to CAD application databases, the question becomes more and more acute: how justified is it to continue using closed, isolated platforms if design data must be available to a wide range of specialists working in dozens of contractors and design organizations?
This vendor-specific technology dependency can significantly limit data management flexibility, slow responses to project changes, and inhibit effective collaboration between participants.
Traditional approaches to data management – including DWH, Data Lake, CDE and PMIS – have focused primarily on storing, structuring and processing information. However, with the development of artificial intelligence and machine learning, there is a growing need for new ways to organize data that not only aggregate, but also identify complex relationships, find hidden patterns, and provide instant access to the most relevant information.
Vector databases – a new type of storage optimized for high-dimensional embeddings – are beginning to play a special role in this direction.
