Dual-scale bionic sensor with high stretchability for AI-enabled squatting motion monitoring.

Area of Science:

• Biomedical Engineering
• Materials Science
• Artificial Intelligence

Background:

• Accurate real-time knee flexion monitoring is vital for sports injury prevention.
• Existing methods face challenges with large flexion amplitudes and complex movement dynamics.

Purpose of the Study:

• To develop a highly sensitive and durable flexible strain sensor for knee flexion measurement.
• To integrate artificial intelligence for precise motion analysis and feedback in sports training.

Main Methods:

• Fabrication of a bionic honeycomb-shaped flexible resistive strain sensor (HSFRSS) with dual-scale structures.
• Characterization of the sensor's performance, including gauge factor, working range, and durability.
• Application of AI algorithms (decision tree regressor and fully-connected neural network) for knee angle prediction and squat pattern classification.

Main Results:

• The HSFRSS demonstrated a high gauge factor (up to 541), wide working range (0-200% strain), and excellent durability (>1000 cycles).
• AI models achieved high accuracy: 99.9% for knee angle prediction and 100% for classifying squat errors.
• The system showed strong generalization across different populations, with prediction accuracy exceeding 97.8%.

Conclusions:

• The developed HSFRSS system offers a promising solution for intelligent motion monitoring in sports.
• The integration of advanced sensors and AI facilitates real-time feedback for posture correction and injury prevention.
• This technology has significant potential for enhancing sports training and performance analysis.