← All Projects
Project
AI powered industrial steel wire rope inspection
For the AI-powered industrial steel wire rope inspection initiative for Raidyn’s AURA system, we developed a comprehensive Computer Vision (CV) and Machine Learning (ML) validation framework.
This project involved deep ML-based algorithmic exploration, utilizing segmentation architectures to enable continuous defect detection across a three-camera setup. To streamline performance tracking, we improved the end-to-end evaluation procedure by unifying model metrics across development, manual, and production environments. Additionally, we implemented a prompt-to-image pipeline using local models on GB10 hardware to efficiently generate synthetic training datasets.
