Pioneering Textile Recycling with Physical AI
7 Aug 2024
The textile industry is facing a growing global crisis, with over 100 million tons of clothing expected to be produced annually by 2025. Yet, shockingly, only 1% of recycled clothing is currently turned into new garments. This staggering statistic highlights the urgent need for innovative solutions to tackle the mounting issue of textile waste. Fortunately, emerging technologies are poised to transform the landscape of textile recycling. Robotic systems equipped with artificial intelligence (AI) and advanced sensors are emerging as game-changers in this space.
Automated Sorting and Identification
One of the key challenges in textile recycling is the accurate sorting, identification and handling of different fiber types, from natural fibers like cotton to synthetic materials like polyester and nylon. This is where cutting-edge technologies are making a significant impact such as near-infrared (NIR) spectroscopy, hyperspectral or multispectral imaging systems for instance, that can accurately detect the fiber composition and presence of contaminants in textile waste.
Disassembly and Component Removal
Apart from sorting, challenges such as scanning of garments for hard elements like buttons and zippers, and then removing them, also currently are awaiting efficient solutions. This process helps prepare the textile portion for more efficient recycling, as it separates the different materials. Beyond textiles, robotic arms are being employed in broader recycling applications, such as sorting and recovering different packaging materials on conveyor lines. This suggests the potential for similar technologies to be adapted for textile recycling, further streamlining the process.
sewts' Brain-in-a-Box
sewts’ brain-in-a-box approach to advanced robotics enables new use cases, positioning it as the go-to technology for complex automation tasks. Our robots rely on high-quality visual data from 2D and 3D vision systems, using this to enhance AI driven tasks emulating human-like cognition. For textile handling, sewts leverages sophisticated material simulations to replicate textile behaviour, aiding in developing smart algorithms and generating synthetic training data to enable rapid adaptation to new requirements.
sewts is building robots capable of complex handling and sorting of textiles based on the development of its “Physical AI”, interconnecting continuous learning from the physical environment and adapting to changes over its long timeframes, evolving and enhancing the efficiency of its tasks.