Development of a Cost-Effective 3D-Printed Smart Glove for Hand and Finger Posture Measurement in Prosthetic Control and Rehabilitation
Disciplines
Artificial Intelligence and Robotics | Biomedical | Signal Processing
Abstract (300 words maximum)
Accurate measurement and prediction of hand and finger posture are essential for advanced prosthetic control and stroke rehabilitation. Commercially available data gloves, such as Manus, CyberGlove, and StretchSense, utilize flex sensors, IMUs, or capacitive sensing to provide high-fidelity motion capture. While these systems offer precise hand tracking, they come with significant limitations, including high cost, software restrictions, and limited adaptability to different hand sizes. The high price of commercial gloves makes them inaccessible for many research labs and rehabilitation centers, restricting their widespread use in personalized healthcare and assistive technology applications.
To address these challenges, this study presents the development of a cost-effective, customizable 3D-printed smart glove designed to provide accurate ground truth measurements for predicting hand and finger postures. The glove integrates low-cost flex sensors and an inertial measurement unit (IMU) to capture finger bending and hand orientation. This design maintains high measurement accuracy while significantly reducing costs compared to commercial alternatives.
To validate the system’s reliability, the glove’s measurements are compared with a Leap Motion Controller, a widely used optical tracking device for hand kinematics. Furthermore, a virtual hand animation developed in Unity is used to visualize gesture predictions, allowing real-time feedback and interactive validation of the system. Electromyography (EMG) signals from the forearm are also recorded to train a deep learning model for accurate static and dynamic gesture prediction, with ongoing efforts to optimize model performance. This system presents a scalable solution for researchers, clinicians, and engineers in the fields of prosthetic control, rehabilitation, assistive technology and human-computer interaction. By providing a low-cost, customizable alternative to expensive commercial gloves, this research advances the development of intelligent wearable technologies that enhance motor function recovery and improve the quality of life for individuals with limb impairments.
Academic department under which the project should be listed
SPCEET - Electrical and Computer Engineering
Primary Investigator (PI) Name
Coskun Tekes
Development of a Cost-Effective 3D-Printed Smart Glove for Hand and Finger Posture Measurement in Prosthetic Control and Rehabilitation
Accurate measurement and prediction of hand and finger posture are essential for advanced prosthetic control and stroke rehabilitation. Commercially available data gloves, such as Manus, CyberGlove, and StretchSense, utilize flex sensors, IMUs, or capacitive sensing to provide high-fidelity motion capture. While these systems offer precise hand tracking, they come with significant limitations, including high cost, software restrictions, and limited adaptability to different hand sizes. The high price of commercial gloves makes them inaccessible for many research labs and rehabilitation centers, restricting their widespread use in personalized healthcare and assistive technology applications.
To address these challenges, this study presents the development of a cost-effective, customizable 3D-printed smart glove designed to provide accurate ground truth measurements for predicting hand and finger postures. The glove integrates low-cost flex sensors and an inertial measurement unit (IMU) to capture finger bending and hand orientation. This design maintains high measurement accuracy while significantly reducing costs compared to commercial alternatives.
To validate the system’s reliability, the glove’s measurements are compared with a Leap Motion Controller, a widely used optical tracking device for hand kinematics. Furthermore, a virtual hand animation developed in Unity is used to visualize gesture predictions, allowing real-time feedback and interactive validation of the system. Electromyography (EMG) signals from the forearm are also recorded to train a deep learning model for accurate static and dynamic gesture prediction, with ongoing efforts to optimize model performance. This system presents a scalable solution for researchers, clinicians, and engineers in the fields of prosthetic control, rehabilitation, assistive technology and human-computer interaction. By providing a low-cost, customizable alternative to expensive commercial gloves, this research advances the development of intelligent wearable technologies that enhance motor function recovery and improve the quality of life for individuals with limb impairments.