Overview
Developed an automated analysis pipeline that processes building geometry from CAD models to evaluate thermal bridging in building envelopes. The tool reduces analysis time from days to hours while improving accuracy through systematic evaluation of design variants.
Process
Started by mapping the manual workflow used by engineers, identifying repetitive tasks suitable for automation. Built Python scripts to extract geometry from Rhino models, generate THERM input files, run simulations in batch mode, and post-process results. Integrated with Grasshopper to enable parametric exploration of design alternatives.
Results
The tool has been deployed on 5 commercial projects, reducing thermal analysis time by 85% and enabling exploration of 10x more design variants. Found previously undetected thermal bridges that would have resulted in 15% higher heating loads.
Reflection
This project taught me the importance of understanding user workflows before automating them. Early versions focused on technical capabilities but missed key usability issues. Iterating with actual users revealed that visualization and reporting were as critical as the analysis itself.