Multi-sensor robotic technology for scrap sorting using RGB vision, HSI, XRT, and LIBS
A scalable multi-sensor platform that improves scrap sorting accuracy by combining RGB vision with hyperspectral imaging (HSI), X-ray transmission (XRT), and laser-induced breakdown spectroscopy (LIBS). The modular design supports starting with minimal sensors and scaling as needed, enabling better identification of complex scrap streams and improved discrimination of visually similar but materially different items for quality control use cases.
Overview
A multi-sensor platform designed to improve identification and sorting of scrap materials. By integrating RGB imaging with advanced sensing—Hyperspectral Imaging (HSI), X-ray Transmission (XRT), and Laser-Induced Breakdown Spectroscopy (LIBS)—the system enhances the accuracy of scrap classification. The modular approach supports cost-effective deployment and scalability, adapting to changing quality control requirements. The technology has demonstrated improved recognition rates and the ability to handle complex material streams.
Technical specifications
- Multi-sensor integration: Combines RGB imaging with HSI for surface chemistry analysis, XRT for density measurement, and LIBS for elemental composition.
- Modular design: Enables starting with a minimal sensor set and scaling up as needed, providing both precision and flexibility in scrap sorting.
- Industrial validation: Demonstrated effectiveness in relevant industrial environments, with higher accuracy in distinguishing visually similar but materially different scrap items.
- Configuration flexibility: Can be configured for specific material categories such as rusted, coated, hazardous, and non-ferrous scraps.
Technology readiness level
The platform is at Technology Readiness Level 6, validated in relevant environments. It is set for further testing, including industrial pilot integration and controlled validation in lab environments.
About Matvision SRL
Matvision SRL is a deep-tech spin-off from the University of Liège that specializes in advanced, robotized sorting solutions for the industrial waste and recycling sectors. The company leverages a decade of research into sensor fusion, artificial intelligence, and high-speed robotics to enable the real-time characterization and separation of complex waste streams, such as shredded metals from end-of-life vehicles and electronic equipment. Its proprietary platform integrates multiple sensing technologies—including X-ray, infrared, and 3D vision—with machine learning algorithms to identify and sort various material types, such as copper, aluminum, and nickel, with high precision in a single pass. This modular approach allows for rapid deployment and customization to specific industrial processing lines. By transforming heterogeneous waste into high-purity secondary raw materials, Matvision helps recyclers capture significant lost value and reduce dependence on virgin resources. The company’s solutions address critical industry needs by increasing recovery rates and enabling efficient, circular supply chains. Beyond manufacturing, Matvision provides consultancy and feasibility studies to assist partners in implementing optimized sorting systems, helping them meet evolving environmental regulations and sustainability goals while improving operational profitability.
Inquire about this partnering listing
Industry applications
delivered to your inbox.