A Big Data Approach to BIM Models

As technology develops at an astounding rate, data has become an intrinsic by-product on major fields, including architecture. Providing designers, contractors and other stakeholders with additional information in order to make informed real time decisions. This paper describes the creation of a series of prototypes which are intended to convert an archive of Building Information Models (BIM) from Little Diversified Architectural Consulting, an early adopter of Autodesk Revit, into a database which can be analyzed using Big Data approaches to data analysis. Using python, as opposed to Revit’s native C# language, provided a more dynamic and simpler pipeline for connecting to the underlying database behind Revit. Use of non-relational databases proved a more efficient system for managing all the information produced by the models. Throughout the use of these tools, we were able to generate our prototypes, which can be divided into three main sections. Starting with altering data inside a Revit model, allowing us to manipulate the databases and add new information as needed. Secondly we collected data between multiple models and stored all the information in a single database. Lastly is visualizing the obtained data from models in order to represent, compare and evaluate the data. After running multiple tests on the models, results proved to reduce repetitive tasks on traditional workflows, while allowing for better analysis and understanding of the implications and processes we use as designers in the 21st century.

https://doi.org/10.35483/ACSA.AM.108.120

Volume Editors

ISBN
978-1-944214-26-5