SpadanaDrawMind
Active Development — AI Model Prototyping & TestingFrom engineering drawingsto engineering intelligence.
We’re developing vision-language AI to connect dimensions, symbols and part geometry — laying the foundation for smarter CAD workflows.
Vision-Language Models · Geometric Reasoning · CAD Integration · Human-in-the-Loop Validation
Current Development Status
SpadanaDrawMind is currently under active development, with software implementation and AI model experimentation underway.
Our current focus is on evaluating AI-based approaches for interpreting mechanical engineering drawings, identifying dimensional and geometric information, and developing reliable workflows for engineering validation.
GPU-accelerated experimentation is an important part of our development process as we evaluate and refine model performance.
Example input drawing · Not an AI result
Schneckenwelle — Thorsten Hartmann · CC BY-SA 3.0 · Unmodified
The challenge
Engineering information, locked in drawings.
Dimensions, symbols and feature relationships must be read and connected manually before they can enter a CAD workflow.
Our approach
Connect what the drawing means.
Our aim is to combine vision-language models with engineering geometry and review, turning drawing interpretation into usable design data.
Development roadmap
Drawing
Dimensions & symbols
AI interpretation
Vision + language
Engineering review
Features & geometry
Editable CAD
Development goal
For mechanical engineers and manufacturing teams.
Our goal: less repetitive data entry, clearer engineering review and easier reuse of drawing information.