Ai-driven reverse engineering for optimal scrap sorting
This solution utilizes machine learning and historical data to enhance scrap sorting precision through AI-driven reverse engineering. It integrates image analysis and chemical tracking to optimize steel production by accurately estimating scrap chemical composition.
Overview
AI-driven reverse engineering for optimal scrap sorting harnesses the power of machine learning and historical production data to improve the precision of scrap classification. By integrating image analysis and chemical composition tracking, this solution enhances decision-making in scrap sorting processes. The technology aims to optimize steel production by accurately estimating the chemical composition of different scrap types, thus improving impurity tracing and operational efficiency.
Technical specifications
- Machine Learning Integration: Utilizes proven ML algorithms for enhanced precision in estimating scrap composition.
- Image Analysis: Links visual features to chemical properties, enabling real-time impurity tracing.
- Chemical Tracking: Tracks and separates tramp elements such as Cu, Sn, and Cr, which can harm steel properties if not managed correctly.
- User-Friendly Application: Provides real-time operator guidance on scrap sorting decisions via a user-friendly app.
- Future Enhancements: Potential for integration with robotics and advanced analytics for improved automation and accuracy.
Technology readiness level
This innovative solution is currently at Technology Readiness Level 6, signifying that it has been demonstrated in a relevant environment through a pilot testing phase. Future steps include refining the model, testing in pilot environments, and evaluating cost-effective sensor options to further enhance classification accuracy.
About University of Belgrade
The University of Belgrade is a comprehensive public research university in Serbia’s capital, uniting a broad network of faculties and affiliated institutes. Industry engages through partner infrastructure such as Science and Technology Park Belgrade—founded with the university, city, and state—which hosts startups and company R&D teams. Clinical translation is supported by affiliation with the University Clinical Centre of Serbia, enabling access for trials and clinician collaborations. A dedicated Center for Technology Transfer provides IP management, licensing, and commercialization support. Membership in European university alliances strengthens cross‑border projects and talent pipelines funded by competitive European and national programs.
delivered to your inbox.