The secondhand garment recycling industry has long faced a core pain point: low sorting efficiency, high labor costs, and inconsistent accuracy. Eton Systems, a member of Swedish textile machinery association TMAS, may be rewriting this rule with its new aUPS module demonstrated at Texprocess in Frankfurt.
The module integrates AI vision recognition with automated material handling, enabling rapid classification of used garments and directly addressing the toughest link in circular textile economy. For secondhand garment traders long reliant on manual sorting, this signals a fundamental shift in cost structure and throughput velocity.
Technology Landing: From Material Handling to Intelligent Sorting
aUPS is not built from scratch; it leverages Eton's widely deployed UPS transport system originally designed for moving garment panels and finished products within factories. By combining this hardware backbone with the proprietary ETONingenious AI platform, the system gains the ability to 'see' and 'decide'.
At the exhibition, the system demonstrated real-time identification and diversion of garments by color, material, and style. The AI was trained on extensive real-world used garment samples, covering common wear, stains, and deformations. This means the system can not only distinguish cotton from polyester but also recognize specific brands or vintage styles—critical for high-value retro garment sorting.
From an industry perspective, the breakthrough lies in compressing the 'manual inspection + manual classification' process into a fully automated line. For European secondhand garment hubs (e.g., in Belgium and the Netherlands), faster sorting directly translates to higher inventory turnover and lower warehousing costs.
Industry Impact: Reshaping the Underlying Logic of Secondhand Garment Supply Chains
The global secondhand garment market continues to expand. According to public industry data, the market exceeded $80 billion in 2025, with annual growth rates above 15%. Sorting has always been the bottleneck: manual sorting handles roughly 200-300 pieces per hour, with fatigue-induced errors.
aUPS module could multiply sorting efficiency several times. More critically, AI sorting enables 'on-demand classification'—dynamically adjusting sorting rules based on downstream buyer requirements (e.g., 'only white cotton T-shirts' or 'reject heavily stained dark outerwear'). This flexibility will reshape bargaining dynamics between exporters and major destinations in Africa, Eastern Europe, and Southeast Asia.
For factories, automation reduces reliance on low-skilled labor. In Europe's rising labor cost environment, this directly alleviates staffing pressures. Higher sorting accuracy also means lower return rates and reduced supply chain waste.