TUM unveils soft exosuit with artificial tendons to help older adults stand and walk

The Technical University of Munich (TUM) on September 28, 2026, unveiled a soft, lightweight exosuit that uses artificial tendons to assist hip extension. It is designed to help older adults stand up from chairs and walk more efficiently. The findings were published in Nature Communications. According to TUM, the device consists of two thigh straps, a waist belt, a small motor unit and two sensors on the thighs, and can be put on or taken off in a few minutes. When it detects that the wearer is starting to stand or walk, the motor pulls the artificial tendon to assist the hip.

TUM says the exosuit supports older adults as they stand and walk and can be worn “over ordinary clothing.” For use in care or rehabilitation, quick donning and removal and minimal discomfort while wearing it are as important as the level of support.

What is new in this announcement is the use of artificial tendons to assist hip extension, combined with machine learning to identify the timing of standing and walking. Exoskeleton research itself follows an established path, but TUM’s device is designed for everyday movement rather than as a rigid frame: It is said to be “put on or taken off in a few minutes” and wearable “over ordinary clothing.” It was tested not on young adults but on 10 older adults ages 69 to 85, with a focus on elder care and rehabilitation. We have not previously reported on this, but the system’s structure is clear from the information provided. Its inputs are movement data from two thigh sensors; machine learning processes and recognizes the movement; and a motor pulls the artificial tendon to provide assistance. The key is not simply that it is an assistive device, but that it synchronizes movement recognition with assistive output in a short time. The study also assessed wearers’ subjective experience of control. A score of more than 6 on a seven-point scale for their sense of agency suggests that the assistance was not overly intrusive, though the trial involved only 10 people. The industry implications may lie in whether lightweight design and recognition accuracy become thresholds for commercialization. Using a garment-like design instead of a frame could help lower adoption barriers in care settings and rehabilitation facilities. For practical use, however, durability, continuous operating time, consistency of fit, and the range of ages and physical abilities covered will be key. TUM says the system was trained initially on movement data from young, healthy people, making the extent to which older adults’ data will be added and how the algorithm will be updated in real-world use the next issues to watch.

The trial involved 10 older adults ages 69 to 85 with reduced lower-limb function who used the exosuit. TUM said they completed an average of 18 sit-to-stand repetitions per minute, two more than before wearing it. The time required for each repetition was reduced by as much as 21.9%, and standing up was described as “easier.” While walking, metabolic cost fell by as much as 27.5%, with an average reduction of 14%. Participants reportedly maintained a sense that they could control their own movements while wearing the assistive device. TUM said it first collected movement data from young, healthy people and used it to train the recognition system and algorithm.

For wearable assistive devices for older adults, the precision of sensors, lightweight actuators and transmission components such as artificial tendons is likely to determine their practicality. Japan has substantial demand for care and rehabilitation applications, but the ability to combine a “garment-like feel” with synchronized movement recognition could become a point of differentiation.