Researchers have developed the prototype of a “comfortable and flexible”soft smart hand exoskeleton,” or robo-glove, which gives information to wearers who need to relearn tasks that require manual dexterity and coordination, for example after suffering a stroke. The current study focused on patients who need to relearn to play the piano as a proof of principle, but the glove can easily be adapted to help relearn other everyday tasks.
Stroke is the leading cause of adult disability in the EU, affecting around 1.1 million people each year. After a stroke, patients often need rehabilitation to relearn how to walk, talk or perform daily tasks. Research has shown that in addition to physiotherapy and occupational therapy, music therapy can help stroke patients recover language and motor function.
But for musically trained people who have suffered a stroke, playing music can itself be a skill that needs to be relearned. Now a study in Frontiers of robotics and AI showed how much soft robotics can help recovering patients relearn music playing and other skills that require dexterity and coordination.
“Here we show that our smart exoskeleton glove, with its touch sensors, flexible actuatorsand artificial intelligence, can effectively help to relearn manual tasks after neurotraumasaid lead author Dr. Maohua Lin, adjunct professor in the Department of Oceanic and Mechanical Engineering at Florida Atlantic University.
Who the glove is suitable for: tailor-made “intelligent hand”
Lin and his colleagues designed and tested an “intelligent hand exoskeleton‘ in the form of a flexible multi-layered 3D printed robot glove, which weighs only 191 g. The entire palm and cuff of the glove are designed to be soft and flexible, and the shape of the glove can be customized to fit the anatomy of each wearer.
Flexible pneumatic actuators at the fingertips generate motion and exert force, mimicking the natural and precise movements of the hand. Each fingertip also contains an array of 16 flexible sensors or “taxels”, which impart tactile sensations to the wearer’s hand when interacting with objects or surfaces. Producing the glove is simple, as all actuators and sensors are put in place through a single molding process.
“When wearing the glove, human users control the movement of each finger to a great extent,” said lead author Dr. Erik Engeberg, a professor in Florida Atlantic University’s Department of Oceanic and Mechanical Engineering.
“The glove is designed to assist and enhance their natural hand movements, allowing them to control the flexion and extension of their fingers. The glove guides the hand, provides support and amplifies dexterity.
The authors predict that patients could eventually wear a pair of these gloves, to help both hands independently regain dexterity, motor skills and sense of coordination.
The AI trained the gauntlet to become a music teacher
The authors used machine learning to successfully teach the glove to “feel” the difference between playing a correct or incorrect version of a beginner piano song. Here, the glove operated autonomously without human intervention, with pre-programmed movements. The song was “Mary had a little lamb”, which requires four fingers to play.
“We have found that the glove can learn to distinguish between correct and incorrect piano playing. That means it could be a valuable tool for the personalized rehabilitation of people who want to relearn how to play music,” Engeberg said.
Now that the proof of principle has been demonstrated, the glove can be programmed to give feedback to the wearer on what went right or wrong in their game, either via haptic feedback, visual or sound cues. These would allow him to understand their performance and make improvements.
Raise the gauntlet for the remaining challenges
Lin added: “Adapting the current design to other rehabilitation tasks beyond music, for example object manipulation, would require customization according to individual needs. This can be facilitated with 3D scanning technology or CT scans to ensure personalized fit and functionality for each user.
“But several challenges in this area need to be overcome. These include improving the accuracy and reliability of touch sensing, improving the adaptability and dexterity of the exoskeleton design, and refining machine learning algorithms to better interpret and respond to user inputs.
This article originally appeared in Frontiers Science News.
