The smart recycling machines in department stores have become an entry point for understanding clothing materials. Recently, a department store in Taiwan introduced a smart clothing recycling machine developed by Peide Sustainability Technology. People only need to place a sweater into it, and the machine can identify blended materials through AI and near-infrared spectroscopy. What surprised many was that a sweater assumed to be wool turned out to be acrylic fiber—essentially plastic. Many people only realized at that moment that many of the soft, warm, or moisture-wicking clothes in their wardrobes are related to petrochemical raw materials.

In 2025, Peide partnered with SOGO Department Store and received more than 12,000 old garments in two days, attracting about 2,000 participants. CEO Liu Yaoda observed that most people do not know that nylon, polyester, and acrylic are all plastics. Only when consumers understand the composition of clothing can they further reflect on the environmental burden of casually discarding it.

However, being able to identify materials does not mean being able to solve Taiwan's old clothing problem. In the past, old clothing recycling in Taiwan centered on "extending wear." Recyclers picked out items in better condition for charity sales, donations, or export. But fast fashion has brought a large number of quickly discarded clothes, and the secondhand market cannot fully absorb them; some overseas markets have also begun restricting imports of secondhand clothes, and the original export model is facing bottlenecks.

What really troubles operators is sorting. If clothes are to be remade into fibers, cotton, wool, nylon, and polyester must be processed separately; otherwise, downstream regeneration plants will find it difficult to obtain raw materials with stable composition and sufficient quantity. At present, the recycling machines at department store sites are limited by size and can only first separate polyester, while the rest still needs to be sent to factories for further sorting. Sorting is only the first hurdle. Even if chemical recycling is used to restore polyester to a material close to virgin quality, cotton, wool, and blended materials may still lack destinations because chemical fiber plants' demand is concentrated in polyester.

Peide therefore invested in physical fiber decomposition, cutting and disassembling clothes into fibers and then spinning them. It took about two years with the team to adjust the process, and several batches of recycled yarn have been produced. It also collaborated with the fashion design team at Shih Chien University to complete more than 20 outfits. However, disassembly shortens fibers and limits strength, so the result may not be suitable for high-performance or skin-friendly clothing. Therefore, reasonable uses must be found according to material quality.

From collection to remanufacturing into yarn, old clothes involve multiple stages: sorting, removing accessories, cutting, fiber decomposition, and spinning. When these are scattered across different sites, costs are higher, and if any link fails to connect, the entire cycle is interrupted. To this end, Peide plans to build a factory in Tainan to concentrate large-scale sorting, crushing, and fiber decomposition, and to connect with nearby yarn factories to form a process of "clothes in, yarn out," attempting to solve the problem of old clothes lacking stable destinations.

There is still a distance from demonstration to mass production. At present, the cost of recycled yarn remains high, and the goal is to control it within 1.5 times that of virgin materials after scaling up; even accessories such as buttons and zippers need to be manually cut off, increasing labor and costs. Liu Yaoda pointed out that in the future it may be possible to cooperate with social welfare groups to provide job opportunities for people with physical and mental disabilities, so that circularity is not only a technical issue; the additional labor and cost burden must also be discussed.

Another supporting measure is a cloud-based record. In the past, after people put clothes into recycling bins, they did not know where they went, and when brands claimed to use recycled fibers, consumers found it difficult to verify. Peide has established a cloud record that logs the recycling point, sorting site, number of items, and material proportions of clothing. If fiber decomposition, spinning, and manufacturing information are connected in the future, it can prove that recycled yarn indeed comes from old clothes recycled at a specific time and site. Liu Yaoda described this as completing the other half of a product's record—no longer only recording how raw materials become clothes, but tracking how old clothes are broken back down into raw materials and put into the next production cycle.

Beyond technology, the biggest obstacle comes from market perception. Many OEM factories wait for brands to take a stance before considering recycled materials, while brands worry about the price and quality of recycled materials. To break this hesitation, it may still be necessary to push from the consumer side to demand action from brands, jointly reversing the market habit in which old clothing circularity costs more.

For Taiwan, a complete chemical fiber and textile supply chain is an advantage for developing "clothes-to-clothes" circularity, but when every link is waiting and watching, it is difficult for new technologies to accumulate output and reduce costs. From recycling and sorting to regeneration, only when every segment connects can circularity truly flow back, rather than stopping at a full recycling bin or the moment of drop-off.

Source: Can wool sweaters be plastic? "AI Clothing Recycling Machine" scans them, and old clothes are sorted, disassembled, and revived｜TVBS GOOD · Originally published on 2026-08-19