For individuals who have endured neurotrauma, for example, a stroke, regular errands can be very difficult in light of diminished coordination and strength in one or both upper appendages. The development of robotic devices to assist in improving their capabilities has been sparked by these issues. However, for more complex tasks like playing a musical instrument, the rigidity of these assistive devices can be problematic.
A first-of-its-sort mechanical glove is loaning a “hand” and giving desire to piano players who have experienced an incapacitating stroke. The soft robotic hand exoskeleton uses artificial intelligence to improve hand dexterity and was developed by researchers from the College of Engineering and Computer Science at Florida Atlantic University.
This robotic glove is the first to combine AI, flexible tactile sensors, and soft actuators into a single hand exoskeleton and “feel” the difference between correct and incorrect versions of the same song.
“Playing the piano requires complex and profoundly talented developments, and relearning undertakings includes the reclamation and retraining of explicit developments or abilities,” said Erik Engeberg, Ph.D., senior creator, a teacher in FAU’s Branch of Sea and Mechanical Designing inside the School of Designing and Software engineering, and an individual from the FAU Community for Complex Frameworks and Cerebrum Sciences and the FAU Stiles-Nicholson Mind Establishment. ” Sensors and soft, flexible materials make up our robotic glove, which helps people relearn and regain their motor skills by providing gentle support.”
The robotic glove was made with special sensor arrays embedded in each fingertip. This new technology, in contrast to previous exoskeletons, provides precise force and direction for recovering the delicate finger movements necessary for piano playing. The robotic glove provides real-time feedback and adjustments by monitoring and responding to users’ movements, making it simpler for them to learn the correct movement techniques.
Researchers programmed the robotic glove to distinguish between the correct and incorrect piano renditions of the well-known song “Mary Had a Little Lamb” in order to demonstrate the glove’s capabilities. To present varieties in the exhibition, they made a pool of 12 unique kinds of mistakes that could happen toward the start or end of a note, or because of timing blunders that were either untimely or postponed, and that continued for 0.1, 0.2 or 0.3 seconds. There were ten distinct song variations, with the correct song playing flawlessly in each of the three groups of three variations.
Random Forest (RF), K-Nearest Neighbor (KNN), and Artificial Neural Network (ANN) algorithms were trained with data from the fingertips’s tactile sensors in order to classify the song variations. The robotic glove was used independently and while a person was wearing it to feel the differences between the correct and incorrect versions of the song. To classify the correct and incorrect song variations with and without a human subject, the precision of these algorithms was compared.
The study’s findings, which were published in the journal Frontiers in Robotics and AI, showed that the ANN algorithm had the highest classification accuracy—97.13 percent with a human subject and 94.60 percent without one. The algorithm was able to identify key presses that were out of time and the percentage error in a particular song with success. These findings highlight the smart robotic glove’s potential to assist disabled individuals in relearning manual dexterity skills like playing musical instruments.
The robotic glove was designed with hydrogel casting and 3D-printed polyvinyl acid stents to integrate five actuators into a single wearable device that fits the user’s hand. Using 3D scanning technology or CT scans, the form factor could be tailored to each patient’s individual anatomy, making the fabrication process novel.
“Our plan is essentially less difficult than most plans as every one of the actuators and sensors are consolidated into a solitary trim interaction,” said Engeberg. ” Importantly, despite the fact that the purpose of this study was to play a song, the method could be used for a wide range of everyday activities and the device could facilitate intricate rehabilitation programs that were tailored to each patient.”
Clinicians could use the data to create customized action plans to find a patient’s weaknesses. These weaknesses might show up as sections of a song that are always played wrong, and they could be used to figure out which motor functions need to be improved. In a game-like progression, the rehabilitation team may prescribe more challenging songs to patients as they progress to provide a customizable path to improvement.
Stella Batalama, Ph.D., dean of the FAU College of Engineering and Computer Science, stated, “The technology developed by professor Engeberg and the research team is truly a gamechanger for individuals with neuromuscular disorders and reduced limb functionality.” Despite the fact that other soft robotic actuators have been utilized for piano playing; our automated glove is the one in particular that has exhibited the ability to ‘feel’ the contrast among right and erroneous variants of a similar tune.”
Maohua Lin, the study’s first author and Ph.D. student, is a co-author; The graduate student Rudy Paul; furthermore, Moaed Abd, Ph.D., a new alumni; all of them are students at the FAU College of Engineering and Computer Science; Boise State University’s James Jones; Darryl Dieujuste, an alumni research colleague, FAU School of Designing and Software engineering; and Harvey Chim, M.D., a professor at the University of Florida in the Division of Plastic and Reconstructive Surgery.
The National Science Foundation, the National Institute of Aging, and the National Institute of Biomedical Imaging and Bioengineering of the National Institutes of Health (NIH) all contributed funding to this study. A seed grant from the FAU Institute for Sensing and Embedded Network Systems Engineering (I-SENSE) and the FAU College of Engineering and Computer Science contributed in part to this study’s funding.