{"id":99,"date":"2020-08-31T12:04:04","date_gmt":"2020-08-31T10:04:04","guid":{"rendered":"http:\/\/matvision.eu\/?p=99"},"modified":"2026-04-09T14:14:10","modified_gmt":"2026-04-09T12:14:10","slug":"design-thinking-building-a-design-system-for-an-existing-product-2","status":"publish","type":"post","link":"https:\/\/matvision.eu\/fr\/design-thinking-building-a-design-system-for-an-existing-product-2\/","title":{"rendered":"Genesis of MATVISION technology at the University of Li\u00e8ge"},"content":{"rendered":"

“It will never happen,” declared the COO of Comet-Traitement back in 2014, referring to the use of robots in recycling. This skepticism was not unfounded. At that time, the recycling industry had never seen robots handling complex tasks like sorting shredded metals. Yet, it takes vision and courage to go against the prevailing sentiment, especially when it comes from an experienced and talented director of a leading company. Our CTO, Godefroid Dislaire, believed that sorting metals with robots was not only possible but also necessary for the future of recycling.<\/p>\n\n\n\n

\n

Sorting metallic waste with robots ? It will never work because of the irregular shape of our scraps.<\/p>\nTop Executive at Comet in 2014<\/cite><\/blockquote>\n\n\n\n

The Birth of a Vision<\/h3>\n\n\n\n

The idea began as a research project at the University of Li\u00e8ge’s GeMMe laboratory. Funded by the Walloon Region under the Reverse Metallurgy<\/a> plan, the project aimed to revive the metallurgy industry by integrating recycled materials into production processes. This ambitious initiative sought to retain highly skilled workers and engineers in the region by shifting focus from traditional metallurgy to recycling.<\/p>\n\n\n\n

The project, known as PICKIT, was a collaboration between three main actors: the University of Li\u00e8ge’s GeMMe lab, Comet Traitements, a company renowned for its innovative approach to recycling, and Citius Engineering, an engineering firm specializing in transitioning lab prototypes to industrial applications. This collaboration laid the foundation for what would become a groundbreaking technology.<\/p>\n\n\n

\n
\"\"<\/figure>\n<\/div>\n\n\n

Overcoming Initial Challenges<\/h3>\n\n\n\n

At the project’s inception, the idea of processing shredded metals with robots was almost unheard of. The primary challenge lay in the irregular shapes and sizes of the metal pieces, which had undergone extensive shredding. In typical industrial applications, robots perform repetitive and simple tasks, handling materials of consistent size and shape<\/strong> with dedicated gripping systems. Adapting the extremely fast delta robots used for packaging chocolates to pick irregular scraps required a significant technological leap. The solution necessitated a high-tech approach, integrating a 3D camera and high-speed computing to determine the optimal picking location based on the volume of each scrap<\/strong> piece.<\/p>\n\n\n\n

With a strong vision and relentless effort, we embarked on building a system capable of addressing these challenges. After finally convincing our industrial partners of our ability to overcome these technical hurdles, they showed marked interest and quickly supported our research project to further develop our robotic solution. The result was the world’s first multi-sensor, multi-output robotic system for processing a mix of non-ferrous metals<\/strong>.<\/p>\n\n\n\n

Technological Breakthrough<\/h3>\n\n\n\n

The PICKIT project combined various cutting-edge technologies to achieve unprecedented accuracy in sorting. Our multi-sensor approach utilized x-ray, hyperspectral imaging, and 3D imaging to provide detailed characterization of the materials. This data fusion significantly enhanced the recognition process. Coupled with multi-output capabilities, the system could sort up to 20 different material families from a mixed input, maximizing the added value for each output.<\/p>\n\n\n\n

The synergy between multi-sensor technology and robotic sorting allowed us to overcome the limitations of existing methods. By 2023, we had constructed the third iteration of our sorting prototype and commissioned a 30-meter-long industrial sorting line, MULTIPICK, equipped with 18 robots, capable of processing 20,000 tons of non-ferrous waste annually.<\/p>\n\n\n\n

Continuous Innovation<\/h2>\n\n\n\n

We are committed to continuously unlocking technological advancements by transferring high TRL (Technology Readiness Level) R&D results from the GeMMe Laboratory at the University of Li\u00e8ge. GeMMe is actively involved in research projects with both local and European key innovators, constantly integrating new sensors and developing advanced AI to process increasingly complex waste streams. As a spin-off from GeMMe, we are dedicated to bringing these technological innovations to the market, leveraging our strong connection with the lab as a significant competitive advantage. This ongoing collaboration ensures that we remain at the forefront of technological advancements, maintaining our edge in the industry.<\/p>\n\n\n\n

\n