A database is a system that integrates, stores, and manages data based on a specific data structure. Similarly, a BIM database is designed to integrate, store, and manage BIM data following a defined data structure. There are several types of BIM databases, including relational BIM databases, object-oriented BIM databases, object-relational BIM databases, and NoSQL (Not Only Structured Query Language) BIM databases. Below is an overview of these BIM database types:
(1) Relational BIM Database
A relational BIM database uses a relational database to store BIM data, relying on the relational model as its foundation. Examples include the IFC database based on SQL Server developed by VTT Building and Transport and ECOMCo., the “BIM Information Integration Platform” created by Dr. Zhang Yang from Tsinghua University, and the ACTIVe3D system by Cruz et al., which uses Oracle for BIM data storage.
The IFC standard, widely used as a neutral and open architectural data expression and exchange format in the engineering and construction industry, includes numerous optional attributes within its entities. However, in practical applications, many component attribute values remain empty, yet relational databases still allocate significant storage space for them, leading to wasted space. Additionally, relational databases have limited scalability, making them less effective at handling the rapid growth and storage demands of BIM data.
(2) Object-Oriented BIM Database
Object-oriented database systems emerge from applying object-oriented programming techniques to database technology. Since object-oriented programming languages are based on objects, developers can store data directly as objects within their programs. Object-oriented databases encompass all the features of relational database systems.
Researchers have explored using object-oriented databases for BIM data storage. For example, Faraj utilized the Object Store database for IFC data, while Lu Ning designed and implemented an IFC database based on the object-oriented Versant Object Database 8. His studies showed that object-oriented IFC databases offer higher efficiency compared to relational databases. However, the theoretical framework for object-oriented databases is still incomplete, and many challenges remain to be addressed.
(3) Object-Relational BIM Database
An object-relational BIM database combines traditional relational database technology with object-oriented methods. This hybrid approach merges the robust functionality and reliability of relational databases with the flexibility and modeling capabilities of object-oriented databases.
Relational databases operate at a lower level, managing data through columns, tables, and rows, while object-oriented databases work at a higher level by encapsulating data within objects. Despite these advantages, object-relational databases are still in the early stages of development and require further research to reach maturity.
(4) NoSQL BIM Database
NoSQL BIM databases are non-relational systems that have evolved rapidly to meet the challenges posed by the explosive growth of the Internet. Compared to traditional relational databases, NoSQL databases offer faster read and write speeds, efficient handling of massive datasets, and better scalability.
NoSQL databases are particularly suited for managing large volumes and diverse types of data, especially in big data applications. They represent a revolutionary shift in database technology, introducing new concepts beyond traditional relational models. However, the theoretical foundation of NoSQL databases remains underdeveloped, lacking comprehensive technical support, and the technology is still considered immature.
Must log in before commenting!
Sign Up