Questions and answers on AI as a tool for the future of textile sorting
Sorting capacity in Europe is only sufficient to deal with a quarter of what is collected today. The rest is incinerated or sent to landfill, despite the fact that as much as 50 to 75% of collected textiles are potentially reusable. Smart technology and automation are opening the door to an efficient and high-tech multi-billion dollar second-hand industry. With the help of artificial intelligence (AI) and machine learning, tens of thousands of different garments can be recognized and those who value the garments will have a good basis for deciding whether the garments should go to second hand or not.
Hi, Susanne Eriksson at Wargön Innovation project manager for AI for circular fashion.
What is the aim of the project?
- In the project, Wargön Innovation, together with other project partners, will put together a world-unique open AI database for intelligent sorting and valuation of second-hand goods.
What is artificial intelligent textile valuation?
- By building up a large database and transferring data from it, robots can be machine-learned to recognize individual garments and their manufacturers. Such robotization has the potential to also assess wear and tear, possible damage and/or odours that affect the value of the garment. In addition, the automated handling is provided with current market data to assess where the demand for the garments is and what the pricing patterns look like to optimize pricing. A greater availability of high-quality second-hand clothing can open doors to more 'slow fashion'. This development is driven by the upcoming revision of the European Textile Waste Directive, which will give us a collection requirement for textiles.
What does the collection requirement that will apply from 1 January 2025 mean?
- This means that all EU Member States must have separate collection systems for textiles. Textile clothing, home and interior textiles, bags and textile accessories are covered. The economic value of second-hand clothing is expected to grow dramatically as collection volumes increase, but for the market to grow, large-scale collection, sorting, valuation, logistics and sales systems are needed.
How much clothing is there to collect?
- In Europe, around 5.4 million tons of clothing, home textiles and other textiles are consumed per year. The average Swede buys 13 kilos of clothes per year and throws away 7.5 kilos in household waste during the same period. 50-75 percent of the volume (depending on where and how it is collected) is possible to reuse for second hand. The problem today is that the capacity to sort the amount of clothes in Europe today is only 1.0-1.5 million tons/year. Today, 3.4 million tons become residual waste or fuel. The average lifespan of a piece of clothing is 2 years and it is used on average 8 times.
What are the economic benefits of textile valuation with AI?
- Second-hand garments account for around 90% of the value of all collected clothes. Other garments have a very low value. Therefore, it is also interesting to increase the accuracy and speed of the assessment of the garments. Automated AI valuation makes it possible to quickly determine the value of each garment among large quantities of clothing. Both the speed and accuracy of the automation have the potential to be far above what human handling can do. The price per garment for automated valuation is significantly lower than for manual work.
What are the health benefits?
- Sorting clothes manually is a monotonous and physically demanding job that can cause repetitive strain injuries. This type of manual assembly line work can be replaced by automated industrial production that requires high production technology skills and can be both developed and operated in Sweden. Parts of the manual work remain in parallel.
What are the environmental benefits?
- Extending the lifetime of clothes threefold, from 2 to 6 years, by reselling them, for example at second hand, reduces greenhouse gas emissions by 70%. The biggest positive environmental impact is in regions outside Europe where most clothes, shoes and textiles are made. These regions currently account for 85% of raw material use, 92% of water consumption, 93% of land use and 76% of greenhouse gas emissions.
80% of a garment's greenhouse gas emissions and 92% of its toxicity come from the early stages of production. The textile industry accounts for 4-8% of global greenhouse gas emissions.