Reverse Engineering Design Logic

Upload a Plan. Get the Concept Back

Architects know how to draw. But how often do we explain why we draw it that way?
What if you could upload a plan. a simple floorplate, a site section, a zoning envelope and get a concept statement back? Not just a description of shapes and rooms, but the strategic logic behind the design.

Core Argument:
Thanks to new AI vision tools and ChatGPT’s reasoning capabilities, we’re entering a phase where the machine can read design not just see it.
This post explores how architects are beginning to reverse engineer their own or others’ drawings to extract the DNA of intent. Not just for fun, but as a serious tool for design critique, learning, and iteration.

Sections:

🔍 1. Seeing Beyond the Lines

AI isn’t just recognizing forms anymore; it’s starting to detect patterns. Upload a plan, and ChatGPT can infer axial hierarchies, zoning strategies, adjacencies, and even programmatic intent with the right prompting.

Example Prompt:

“Describe the organizing principle of this floor plan. Identify circulation logic, key zones, and possible conceptual drivers.”