Robotic jacket to treat paralysis

History

2025: Product Announcement

Researchers at Harvard University have developed a soft robotic jacket that can restore motor function in people with upper limb motor impairment after a stroke or amyotrophic lateral sclerosis. The device uses machine learning technology and a network of smart sensors to personalize patient movement support. The results of the study are published in the scientific journal Nature Communications in August 2025.

According to Interested Engineering, the development was carried out by specialists from the Harvard John Paulson School of Engineering and Applied Sciences together with doctors from Massachusetts General Hospital and Harvard Medical School. The robotic jacket is designed to help with basic activities on a daily basis, including eating and drinking.

𝓲

Фото: Harvard University

The device looks like a regular vest, but contains a complex system of smart sensors and micro-drives under the fabric. Machine learning algorithms analyze the slightest attempts at motion that cannot be seen with the naked eye and amplify them to perform full-fledged motor actions.

The robotic system learns the individual features of each user, adapting to the unique characteristics of his muscle activity. The researchers integrated a machine learning model with a physical hysteresis model to increase control transparency and adapt to practical use in real-world settings. Clinical trials were conducted with five patients after stroke and four people with amyotrophic lateral sclerosis. The device controller detects shoulder movements with an accuracy of 94.2% with minimal changes in the angles of the shoulder joint - a rise of 3.4 degrees and a drop of 1.7 degrees.

The robotic jacket reduces the force required to lower the hand by 31.9% compared to the base controller. The device improves the quality of movements by increasing the amplitude of lifting and lowering the shoulders by 17.5 degrees, flexing the elbow by 10.6 degrees and flexing-extending the wrist by 7.6 degrees.^[1]

Notes