Development of Soft Robotic Hand Data Glove for Rehabilitation and Gesture Recognition
Disciplines
Applied Mechanics | Biomechanical Engineering | Biomedical | Robotics
Abstract (300 words maximum)
Hand rehabilitation is a critical component in the recovery process for individuals who suffer from stroke, traumatic injuries, or neurological disorders that can impair hand function and dexterity. Traditional rehabilitation methods often involve repetitive exercises and protocols that can be tedious and discouraging for patients, potentially preventing their progress and engagement to the rehabilitation program. The integration of soft robotic technology with 3D printing offers a promising solution to enhance hand rehabilitation as well as gesture recognition for prosthetic control. A soft robotic based hand glove, which is a wearable device made from compliant and flexible materials, can provide dynamic assistance and support to the hand during rehabilitation exercises, while also enabling accurate tracking of hand and finger movements for gesture recognition applications. Utilizing 3D printing technology, soft robotic hand gloves can be customized to fit individual hand sizes and shapes, ensuring a comfortable and personalized fit. In addition, the flexibility of 3D printing allows for the incorporation of various sensors and actuators within the glove's structure, enabling precise monitoring of hand and finger motions as well as assistance during rehabilitation exercises. For this purpose, we manufactured a soft robotic hand glove using both flexible and rigid materials as well as a compliant structure for ease of motion and for sensor accuracy. The glove consists of 10 integrated flex sensors equipped to measure simultaneous angle position data from each finger. A microcontroller-based data acquisition system is developed to collect sensor data and compute finger positions. This data is used to control a Matlab Simscape based hand model animation. The developed 3D printed hand glove can be used to collect ground truth together with EMG/Ultrasound sensor data in deep learning-based model development for hand/finger motion prediction.
Academic department under which the project should be listed
SPCEET - Electrical and Computer Engineering
Primary Investigator (PI) Name
Dr. Coskun Tekes
Development of Soft Robotic Hand Data Glove for Rehabilitation and Gesture Recognition
Hand rehabilitation is a critical component in the recovery process for individuals who suffer from stroke, traumatic injuries, or neurological disorders that can impair hand function and dexterity. Traditional rehabilitation methods often involve repetitive exercises and protocols that can be tedious and discouraging for patients, potentially preventing their progress and engagement to the rehabilitation program. The integration of soft robotic technology with 3D printing offers a promising solution to enhance hand rehabilitation as well as gesture recognition for prosthetic control. A soft robotic based hand glove, which is a wearable device made from compliant and flexible materials, can provide dynamic assistance and support to the hand during rehabilitation exercises, while also enabling accurate tracking of hand and finger movements for gesture recognition applications. Utilizing 3D printing technology, soft robotic hand gloves can be customized to fit individual hand sizes and shapes, ensuring a comfortable and personalized fit. In addition, the flexibility of 3D printing allows for the incorporation of various sensors and actuators within the glove's structure, enabling precise monitoring of hand and finger motions as well as assistance during rehabilitation exercises. For this purpose, we manufactured a soft robotic hand glove using both flexible and rigid materials as well as a compliant structure for ease of motion and for sensor accuracy. The glove consists of 10 integrated flex sensors equipped to measure simultaneous angle position data from each finger. A microcontroller-based data acquisition system is developed to collect sensor data and compute finger positions. This data is used to control a Matlab Simscape based hand model animation. The developed 3D printed hand glove can be used to collect ground truth together with EMG/Ultrasound sensor data in deep learning-based model development for hand/finger motion prediction.