ShilpaSutra: An Innovative AI-Driven Platform for Natural Language to CAD/CFD Engineering Design

ShilpaSutra is an open-source AI-driven platform that bridges natural language input to CAD/CFD engineering design. Its name (from Sanskrit "Shilpa" = craft, "Sutra" = formula) reflects its mission to "democratize engineering design"—enabling users to express design intent via natural language, with AI automating modeling and simulation. It doesn't replace existing CAD/CAE tools but acts as an intelligent bridge between human creativity and professional tools, lowering input barriers while ensuring the professionalism of the output.

Traditional CAD/CAE tools require expertise in complex interfaces and geometric modeling, limiting efficiency and innovation. Many professionals with engineering intuition are excluded due to software learning curves, and impromptu ideas are often put aside due to high modeling costs. ShilpaSutra addresses this by enabling natural language-driven design, automating the transition from intent to digital models and simulations.

ShilpaSutra's core modules include:

1. Dialogue CAD Designer: AI chat sidebar for natural language interactions (e.g., "create a 50mm diameter cylinder") to automate modeling, with a layout similar to mainstream CAD tools.
2. Text-to-CAD Conversion: LLM-powered API (/api/generate-cad) that parses natural language to generate parameterized 3D models (supports 10+ export formats like STEP/STL).
3. CFD/FEA Simulator: Automated workflow (geometry prep, AI-driven meshing, natural language working condition setup) using OpenFOAM (CFD) and CalculiX (FEA).
4. Assembly & Parts Library: Supports component constraints, BOM generation, and a searchable parameterized parts library.
5. Engineering Drawings: Auto-generates 2D drawings with GD&T and size annotations.

ShilpaSutra uses modern web tech:

• Frontend: Next.js14 (App Router), TypeScript, Tailwind CSS, Three.js + React Three Fiber (3D rendering).
• State Management: Zustand for efficient data sync.
• UI: Radix UI (headless components), Framer Motion (animations), Lucide Icons.
• Backend Plans: Integrate OpenFOAM (CFD), CalculiX (FEA), CadQuery/Build123d (geometric kernel).
• Deployment: Vercel (Serverless, global CDN).

AI is integrated throughout ShilpaSutra's flow:

• Intent Understanding: LLMs (Claude/OpenAI) parse natural language to extract design parameters and constraints.
• Knowledge Reasoning: Provides engineering suggestions (e.g., material/wall thickness recommendations for load-bearing parts).
• Parameter Optimization: Closed-loop design-simulation-optimization to meet performance goals.
• Error Detection: Identifies potential issues (interference, insufficient wall thickness) before they occur.

ShilpaSutra applies to:

• Rapid Concept Validation: Generate 3D models quickly for design reviews/market validation.
• Design Automation: Fast variant design of series/custom products via parameterized templates.
• Cross-Discipline Collaboration: Non-CAD professionals (sales/marketing) contribute via natural language.
• Education: Lowers CAD/CAE learning curves for students to focus on design principles.

Planned features include:

• Deepen Three.js integration in Designer/Simulator pages.
• Connect Claude/OpenAI backend for Text-to-CAD.
• Real-time CFD via WebSocket + OpenFOAM.
• FEA with CalculiX integration.
• CadQuery/Build123d geometric kernel.
• Real-time collaboration (WebSocket), Git-based version control, file import (STEP/STL/OBJ), multi-user workspaces.

ShilpaSutra represents a frontier in AI-CAD/CAE fusion, showing how natural language can lower tool barriers and AI can boost design intelligence. While still in early stages, its vision (cloud-based, intelligent, collaborative) aligns with next-gen engineering tools. As LLMs and engineering knowledge advance, it has the potential to become a standard part of engineers' workflows.