KAIST and Stanford unveil clothing that climbs onto its wearer

Researchers from KAIST and Stanford University have developed SWAG (Self-Wearing Adaptive Garments), a robotic clothing system that uses soft growing robots to dress people with limited mobility. The garments deploy themselves by inflating internal tubes that turn the fabric inside-out, reducing friction and eliminating the need for cameras or precise planning. The technology, which won the Best Paper Award at ICRA 2025, aims to assist elderly and disabled individuals who struggle with dressing.

Getting dressed is the sort of task nobody thinks about until they can no longer manage it. For millions of older and disabled people, pulling on a sleeve is a daily negotiation with buttons, shoulders, and gravity, and it usually needs a second pair of hands. A research team from South Korea and the United States would rather the garment did the reaching. Scientists from KAIST in Daejeon and Stanford University have unveiled a wearable system that climbs onto its owner, with none of the robot carers https://thenextweb.com/news/how-humane-is-the-uks-plan-to-introduce-robot-companions-in-care-homes-syndication or wearable exoskeletons https://thenextweb.com/news/ai-exoskeleton-wearable-german-bionic-startup that tend to headline the assistive-tech beat, as first reported by Reuters https://www.reuters.com/world/asia-pacific/south-korea-us-team-unveils-robotic-technology-that-dresses-wearer-2026-07-17/ . They call it SWAG, for Self-Wearing Adaptive Garments, and the acronym is only half a joke. Instead of a rigid arm that grips a jacket and wrestles it over a body, the clothing deploys itself. The trick is borrowed from vine robots, the soft growing machines developed in Allison Okamura’s lab https://charm.stanford.edu/Main/AllisonOkamura/ at Stanford. Those devices creep forward by everting from their tip, the way a party horn unrolls, and have been used to snake through rubble and inside the body. SWAG applies the same idea to clothing. Thin internal tubes, which the team calls subvines, inflate and extend along the limb, turning the fabric inside-out as they advance. That eversion is the clever bit. Because the material unrolls onto the skin rather than dragging across it, friction stays low, which matters a great deal when the skin in question is thin, bruised, or sore. Crucially, the design never pressurises the whole outer layer. The channels sit around a cylindrical fabric sheath and do the pushing on their own, so the garment stays soft while it works. The project pairs KAIST’s https://www.kaist.ac.kr/en/ Jee-Hwan Ryu, who studies soft growing robots, with Okamura, a veteran of the field. “Our work extends the core principles of soft growing robots into assistive dressing technologies that can directly improve daily life,” Ryu said. The lead author is Nam Gyun Kim, a KAIST researcher, working alongside Stanford mechanical engineering students on the hardware. The prototypes so far include sleeves, jackets, and trousers, each built around the same unfurling mechanism, and each pitched at people with limited mobility or those recovering from injury. The appeal is easy to grasp for anyone who has watched a relative stall at the wardrobe. Conventional dressing robots lean on cameras and careful planning to follow a moving, unpredictable body, and they can press too hard or lose the thread entirely. SWAG sidesteps much of that. The garment conforms to whatever posture the wearer happens to hold, so it does not need a flawless picture of the arm before it starts. The field has taken notice. The paper ran in IEEE Robotics and Automation Letters in November 2025, and this year it took the journal’s Best Paper Award https://www.eurekalert.org/news-releases/1132958 at the ICRA conference, one of five chosen from more than 1,700 published across the year. It was the second year running that Ryu’s group claimed the honour, a detail that hints at how quickly soft growing robots have moved from curiosity to credible tool. This being research, the caveats are real. The garments remain prototypes, and performance leans on the fabric and on the tradeoff between stiffness and flexibility, which grows thornier as more subvines are added. Dressing is not a solved problem in isolation, either. The wider drive to automate care stretches from exoskeletons with a sense of muscle memory https://thenextweb.com/news/new-exoskeleton-for-humans-comes-with-ai-brain-muscle-memory to humanoids courting billion-dollar valuations https://thenextweb.com/news/limx-dynamics-2-2-billion-humanoid-pre-ipo , and each meets the same test of trust around vulnerable people. Even so, there is something disarming about a machine that does not loom. A jacket that quietly walks itself up your arm is a gentler proposition than a robot leaning over the bed. Whether SWAG ever reaches a real wardrobe is another matter, and one the team has not settled. For now it is a demonstration, a small proof that clothes can, in their modest way, learn to dress you. Get the TNW newsletter Get the most important tech news in your inbox each week.