AI Automation for Architectural Casework Drawings

Job ID: 40371487

Budget: $5,000 – $10,000 USD

AI Tool: Convert Architectural Casework Elevations (PDF) into Editable AutoCAD DWG Files with Self-Training Interface
Project Description
I need a custom AI tool that automatically reads architectural casework elevation drawings (PDFs) and generates complete, editable AutoCAD DWG files using my existing multiple block libraries.
The tool must detect cabinets, sinks, fixtures, shelving, countertops, and miscellaneous equipment, then place the correct blocks with accurate dimensions, countertop outlines, and section views.
It must also include a natural language self-training interface so a non-technical user can teach and refine the AI using plain English rules or by correcting the output drawing.
This is Phase 1 (Proof of Concept) of a larger initiative. Future phases will expand to multiple elevations and additional manufacturers.
Sample Files Provided

Before: Architectural elevation A407 (page 26)
After: Mott shop drawing 2-08 (page 30)
Full block library, spec sheets, and catalogs will be supplied

Key Requirements
Input

PDF casework elevation drawings (sometimes original DWG)
My organized AutoCAD block library
Project-specific catalogs and reference documents

Output

Clean, editable AutoCAD DWG file containing:
Correctly placed blocks for all detected casework, sinks, fixtures, shelving, etc.
Accurate dimensions (height, width, spacing)
Countertop outlines with detailed top-view information
Section views based on cabinet types
Proper layers, line types, and drafting standards
Visual flags / notes for any unmatched or uncertain items


Core Features

YOLO-based (or equivalent) object detection for cabinets, sinks, pegboards, etc.
Vision LLM assistance for annotation and context understanding
Block matching engine with fallback logic
Natural language rule engine (“teach” the AI in plain English)
Self-training interface (view, edit, enable/disable rules)
Drawing-based learning (compare AI DWG vs. user-corrected DWG)
Confidence warnings and visual flagging of uncertain items

Technical Preferences

Detection: YOLOv8 / YOLOv11 (fine-tuned on my samples) + Vision LLM (GPT-4o or Claude)
DWG generation: ezdxf (preferred for standalone) or pyautocad
Rule storage: Human-readable JSON/YAML
Training interface: Lightweight GUI (Tkinter / PyQt) or simple web interface

Deliverables

Fully working Python tool (script + optional GUI)
Natural language rule trainer and visual rule manager
Clean DWG output meeting my drafting standards
Full source code, trained model weights, training scripts, and documentation
Setup guide and user manual
Log/report of unmatched or low-confidence items

Timeline
POC Phase 1 (single elevation) within 2–3 weeks
Full Phase 1 within 6–8 weeks (flexible)

Budget
Please provide your fixed price for the POC 1 & POC 2 (single elevation with full DWG output + basic self-training interface). Higher budget available for excellent quality and clean code.
How to Apply
Please reply with:

Your proposed technical approach (tools and libraries)
Estimated timeline for the POC 1
Fixed price for the POC 1
Links or screenshots of any similar past projects (CAD/DXF automation, technical drawing conversion, or vision AI on drawings)

I will provide all sample files immediately upon hiring.
Looking forward to your proposals.
Related categories: Python Machine Learning (ML) Computer Vision YOLO