A Proposal of Bioinspired Soft Active Hand Prosthesis.

Area of Science:

• Robotics
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Traditional rigid prostheses limit natural interaction and functionality.
• Soft robotics offers adaptive and compliant solutions for human-robot interaction.
• Bioinspiration provides a design framework for advanced prosthetic devices.

Purpose of the Study:

• To develop a dexterous, bioinspired soft active hand prosthesis emulating human hand biomechanics.
• To integrate a myoelectric control system for intuitive gesture recognition.
• To validate the prosthesis's grasping performance with various objects.

Main Methods:

• Designing and 3D printing a skeletal structure based on human hand anatomy.
• Utilizing soft materials for musculoskeletal components to ensure compliance.
• Developing a recurrent neural network (RNN) for electromyography (EMG) signal classification.
• Testing grasping capabilities with objects of diverse sizes, textures, and shapes.

Main Results:

• The prosthesis successfully emulated biomechanical properties, enabling natural design and improved grips.
• The RNN model achieved 87% accuracy in real-time hand gesture classification.
• The soft hand prosthesis demonstrated adaptability, soft grasping, and mechanical compliance with daily objects.

Conclusions:

• The bioinspired soft active hand prosthesis offers a promising solution for restoring or enhancing hand capabilities.
• The integrated myoelectric control system provides intuitive and accurate gesture recognition.
• The developed prosthesis exhibits excellent adaptability and compliance for real-world object manipulation.