Autodesk University 2024 The design and Make Conference
Date:
Introduced Silman Colab, Computational design Research Group at Silman TYLin and my current work on Optimizing a Structural Truss System for Cross-section and Embodied Carbon Emissions using Genetic Algortihms and later converting the meta-data to represent it as a Graph. This unique research represents a BIM model as Graph.
#AutodeskUniversity2024_NoFluff_Presentation_Summaries
A blend of Revit API (C#, PyRevit), and Autodesk Platform Services helps optimize structures and reduce carbon emissions.
📌 Key Takeaways: • Integrate Large Language Models with Revit and Autodesk Platform Services for smarter structural optimization. • Use graph databases like Neo4j for data handling. • Combine AI and programming to enhance structural engineering.
💡 Practical tips: → Prepare BIM data using PyRevit and C# add-ins. → Query Revit models with Autodesk Platform Services and GraphQL for precise data retrieval. → Enhance AI models using knowledge graphs and Retrieval Augmented Generation (RAG).
🛠️ Tools and technologies used: • Revit and Revit API • PyRevit • C#/.NET • Autodesk Platform Services (APS) • GraphQL • Neo4j (graph database) • Langchain (AI framework) • Karamba (Grasshopper plugin) • Rhino.Inside Revit
How Revit (and API) was used: Revit served as the foundation for structural models optimized using AI. Custom add-ins extracted data and operated external databases.
How AI was used: LLMs optimized structural models and enhanced data querying.