Web-Based IFC Application for Infrastructure Models: Enhancing User Interaction Through Natural Language Processing

Building Information Modeling (BIM) has revolutionized the Architecture, Engineering, Construction, and Operations (AECO) industry by providing a more collaborative approach to planning, designing, and managing building projects. Among the various standards used in BIM, Industry Foundation Classes (IFC) stand out as a vital open BIM standard by promoting interoperability across different software applications and enabling continuous data exchange throughout a project's lifecycle. As infrastructure projects grow in complexity, more sophisticated tools are being developed to improve model visualization and user interaction, especially in web-based environments. This thesis addresses this need by developing and evaluating a web-based IFC viewer adapted for infrastructure models. By leveraging cutting-edge open-source JavaScript libraries and Natural Language Processing (NLP) technologies, the proposed solution enhances visualization and user interaction within BIM workflows. The innovative tool offers synchronized views of model alignments, profiles, and cross-sectional visualizations at specific points along the 3D alignments. As a result, it provides users with a more intuitive and broad understanding of the model. A distinctive feature of the application is an integrated Conversational Agent which allows users to interact with the model in a more accessible and user-friendly approach. This functionality is delivered through a web interface, eliminating the need for installed software and local processing power, as a result making it accessible from any device with an internet connection. A real-world case study is presented which introduces a workflow that starts with a BIM authoring software model, extends to adding alignments, and concludes with exporting the models to IFC format for web visualization. This workflow was used to demonstrate a practical application that can be replicated with other models. The model is then used to test the IFC Viewer and the integrated Conversational Agent, analyzing the advantages and the drawbacks that arise when using these tools. The results describe the innovative solutions this viewer brings and also the potential future improvements that can take it a step further. Accessing advanced visualization with a user-friendly interface can improve how we manage and interact with digital infrastructure models. The developed IFC viewer makes it easier to work with complex infraBIM data, but it also introduces innovative tools that can be further explored in future research within the AECO industry.