Robotics and Computer Vision in Rare Earth Waste Management: An Essential Combination26 October 2023

Rare earth elements, strategic resources and key invisible components of the technological revolution since the late 20th century, have become ubiquitous in our lives. Powering our smartphones, electric vehicles, data centers, and other everyday technologies, they have now become indispensable for the proper functioning of the global economic machinery.

The technological revolution, coupled with geopolitical and economic concerns, has generated an insatiable demand for rare earths. However, this relentless race for deposits and extraction has not been without consequences. Besides the pollution stemming from mining operations, we have witnessed a sharp increase in electronic waste for several decades. According to the United Nations, the quantity of electronic waste doubled between 2005 and 2019, reaching 53.6 million tonnes, with alarming projections of 74 million tonnes by 2030. Meanwhile, only 20% of this waste is currently treated and recycled, representing a significant waste, especially when considering the financial, energy, and environmental costs associated.

This realization has prompted numerous stakeholders, both from the institutional (e.g., European Parliament, etc.), economic, associative, and civil sectors, to rethink our economic model, aiming for a more responsible utilization of rare earths throughout their entire life cycle. This reorientation will not be simple. Beyond the paradigm shift it entails, it also faces significant operational challenges. Indeed, the complexity of managing rare earths in their recycling lies in their often scattered presence within a multitude of products and the technical difficulties associated with their extraction, separation, and recycling. These elements are frequently mixed with other materials, making their recovery even more challenging. Their treatment is also often hazardous and poses significant health risks due to the presence of toxic substances, thereby increasing costs due to the advanced technology required for these operations.

In the face of these challenges, robotics naturally emerges as a formidable ally. Specially designed robots can be programmed to efficiently disassemble, sort, and recycle end-of-life electronic components, reducing human exposure to hazardous chemicals. However, when combined with computer vision technologies, their full potential is realized.

Computer vision technology is a subcategory of artificial intelligence (AI) and machine learning designed to enable computers to “see” and understand the visual world. Using cameras, it collects images, preprocesses them to make them intelligible, identifies key features of these images (e.g., color, texture, shape, etc.), interprets them to recognize objects, and makes decisions accordingly. This ability to comprehend its environment allows computer vision to answer questions such as “What is in this image?” or “What action is taking place in this video?”. It is used across various sectors, from autonomous vehicles that use computer vision to detect hazards, to security surveillance to identify suspicious behaviors, to automatic detection of falls in nursing homes, and to estimate foot traffic in shopping centers.

In the case of (re)valorizing rare earths found in our waste, computer vision indeed enables robots to “see” and interpret their visual environment with great precision, while minimizing errors and risks associated with rare earth management. This technology is already successfully employed in numerous waste management facilities in Europe, notably through Recycleye Vision (a British startup), which uses computer vision to scan and identify mixed waste. Each waste element is classified, optimizing machine operations and providing precise composition data. Furthermore, once waste is identified, sorting robots can intervene to automate material sorting, which foster the recycling of various electronic waste components.

The fusion of robotics and computer vision provides a glimmer of hope by reducing human and environmental risks, improving waste management efficiency, and offering the opportunity to reintegrate these raw materials into the production cycle. However, it is essential to recognize that these technologies offer only a partial response to the consequences of an outdated linear economic model. We must rethink our approach by anticipating these issues from the product design stage, including circularity-by-design principles.

Only a new circular economic system, where products are designed to last and generate minimal or no waste, will sustainably address the problem of rare earth waste. As we embark on this transformation, robotics and computer vision play a central role, but they are just one piece of the puzzle for a more sustainable future.