Adaptive, sensor-laden garments could provide a new way for quadriplegics to control their wheelchairs. The system, which is still in an early stage of development, identifies the ideal set of movements that can be employed as control commands for each individual user. “We think this will benefit the most difficult patients, such as those who can move only their head or shoulders,” says Alon Fishbach, a scientist at Northwestern who is among those developing the device.
People with high-level spinal-cord injuries often lose control of their hands, but they may still be able to move their shoulders or chests. More and more such patients survive their injuries, thanks to respiratory devices that help them breathe. But these people have limited options when selecting devices to control their wheelchairs or computers. They might use a sip/puff switch, which converts the user’s sip or puff of air into a specific command, or a headswitch, which records head movements via a switch on the back of the wheelchair. “But the disadvantage of these devices is that patients must fit the capacities of the machine, rather than the other way around,” says Ferdinando Mussa-Ivaldi, another Northwestern scientist working on the device. “If a patient can move their right side more than their left, an intelligent interface could pick up on this.” Mussa-Ivaldi directs the Robotics Laboratory of the Rehabilitation Institute of Chicago, where the research took place.
To overcome this design flaw, the researchers are developing an adaptive device using sensor-laden fabric. The garment is printed with 52 flexible, piezoresistive sensors developed at the University of Pisa. These sensors are made of electroactive polymers that change voltage depending on the angle at which they are stretched. The sensors can detect fine scale movements of the upper body and arms.
While the concept is similar to sensor arrays that animators use to convert real-life movement into animation, the device has a unique analysis component. A specialized statistical algorithm analyzes the signals recorded from each sensor to detect the specific movements that can be most reliably recorded from that individual. The algorithm also determines the maximum number of distinguishable motions that each person can make, which in turn dictates the number of commands that person can use.
To “train” the device, researchers ask users to move naturally. The device then selects specific movements that can be used to control the velocity, acceleration, or direction of the wheelchair. A virtual-reality training session helps the patient fine-tune his or her ability to pilot the wheelchair and offers feedback on his or her performance.
The researchers have so far tested the garment on one patient, who was able to successfully move around a virtual course. Next they plan to try out the device with patients who are more severely disabled. (The first patient still had use of his hands.)
Ultimately, Mussa-Ivaldi and his colleagues plan to develop a device that can adapt not only to the individual capabilities of the patient, but also to an individual’s changing ability. “Residual movement might change over time,” says Daofen Chen, program director of Systems and Cognitive Neuroscience at the National Institute of Neurological Disorders and Stroke in Bethesda, MD. “Many young subjects have great potential in brain plasticity. Over time, they will either have some recovery or they will tap into an uninjured part of the brain.” Patients with some degenerative diseases, such as amyotrophic lateral sclerosis, show the opposite pattern, with worsening paralysis over time.
The researchers are currently focusing on a system to control wheelchairs, but they say the device could be used to control a wide range of machines. This may actually be the best use for such a system, says Steven Edwards, a columnist and quadriplegic who currently uses a sip/puff switch to control his wheelchair. “If they could make a device where shoulder controls are linked to a lot of different things, that would be useful for environmental control, such as the TV,” says Edwards. “A video game controller would be great … My brother is a big gaming addict, and I can’t compete with him anymore … It would be cool to be able to do.”
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
Data analytics reveal real business value
Sophisticated analytics tools mine insights from data, optimizing operational processes across the enterprise.
Driving companywide efficiencies with AI
Advanced AI and ML capabilities revolutionize how administrative and operations tasks are done.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.