System Overview

WrenchML is a powerful platform that converts CAD files into synthetic datasets for training computer vision models designed to detect manufacturing defects. This document provides a high-level overview of the system, its components, and its capabilities.

What is WrenchML?

WrenchML is a specialized tool that bridges the gap between CAD (Computer-Aided Design) data and AI model development for manufacturing quality control. By leveraging your existing CAD files, WrenchML generates photorealistic datasets with simulated defects, allowing you to train robust vision models without the need for extensive real-world data collection or defect examples.

Key Components

The WrenchML platform consists of several integrated components:

• 3D Model Conversion Pipeline: Transforms your STEP/CAD files into optimized 3D formats suitable for rendering.
• Defect Simulation Engine: Applies realistic manufacturing defects to your 3D models according to your specifications.
• Synthetic Data Generator: Creates photorealistic images of your parts with and without defects in various environments.
• Dataset Management System: Organizes generated images into structured datasets for model training.
• Export Interface: Provides your data in formats ready for use with popular computer vision frameworks.

How It Works

1. Upload: You upload your CAD file (STEP format) to the platform.
2. Convert: WrenchML converts your CAD file into a GLB format suitable for rendering.
3. Configure: You specify the defect types, materials, and scene parameters you want to simulate.
4. Generate: The system renders photorealistic images of your part with the specified defects.
5. Export: The resulting dataset can be exported for use in training your computer vision models.

Benefits

• No ML Expertise Required: You define the task, WrenchML delivers a deployable model.