TOMRA Mining's OBTAIN™ transforms mineral classification through deep learning innovation

Artificial intelligence (AI) is revolutionizing mineral classification, propelling the industry into a new era of efficiency and accuracy. Stefan Jürgensen, head of software team at TOMRA Mining, explains how sensor-based sorting has evolved with AI.

For decades, AI has quietly enhanced computer systems' ability to mimic human decision-making processes, significantly benefiting solutions like TOMRA's sensor-based sorting. These systems not only automate the classification process but also enhance its precision and efficiency, delivering value to mining operations.

TOMRA Mining's experience with AI dates back to 1993 when its predecessor, CommoDaS, developed sensor-based sorters using AI in image processing systems to identify particle properties. Over the years, sensor-based sorting technology has evolved, with TOMRA leveraging machine learning in its X-ray transmission (XRT) and NIR sorters for the past decade.

Now, the company is ushering in a new era of classification with its latest innovation, OBTAIN™, harnessing Deep Learning to provide granulometric precision in high-performance particle classification. This revolutionary solution enhances process capacity and quality, elevating material recovery to new heights with detailed and precise data for informed decision-making.

AI encompasses two subfields that have seen significant advancements: Machine Learning, which recognizes patterns, learns from data, and improves without explicit programming, and deep learning, which employs artificial neural networks to analyze data and solve complex problems. Both technologies process vast amounts of data rapidly, making autonomous decision-making possible.

Machine Learning and Deep Learning can further enhance the classification process for mining operations already utilizing sensor-based sorting, while also unlocking new opportunities by enabling the treatment of low-grade materials previously considered waste. Another advantage of AI is the wealth of data it generates and processes, allowing mining operations to glean valuable insights into sorter performance and improve preventive maintenance.

It's worth noting that Machine Learning and Deep Learning are not "one-size-fits-all" solutions. Leveraging their extensive experience and intensive R&D activities in AI, TOMRA will always select the most suitable technology for each case.

The OBTAIN™ Revolution: Unprecedented Production in Sensor-Based Sorting

This revolutionary software employs a neural network to precisely identify the properties of each particle, regardless of the machine's capacity. This achieves unparalleled rigor and reliability in material detection and ejection, even for undesired materials.

As a result, mining operations now have greater flexibility to increase sorter production while maintaining consistency and efficiency or to enhance classification precision without sacrificing performance. 

This technology, offers advantages to fully operational mines, improving recovery rates and enhancing product quality while maintaining existing production levels. For mines with additional capacity, it facilitates increased production without compromising product quality. Additionally, this innovative technology allows mines to valorize materials that were previously considered unexploitable, such as low-grade ore, mineral waste deposits, or materials previously untreated due to perceived lack of profitability.

Furthermore, OBTAIN™ adds value to mining operations by providing extensive, detailed, and precise data, such as precise granulometric distribution of incoming material. When used in conjunction with TOMRA Insight, this information provides customers with detailed reports on sorter and component performance, optimizing processes and facilitating predictive maintenance planning.

The OBTAIN™ software has been developed for TOMRA's XRT sorters and will be available in new models, with an upgrade package also provided for machines already in operation. This presents customers with operational TOMRA XRT sorters with a superb opportunity to significantly enhance classification, provided it is the appropriate solution for their specific case.