Development of an Environment-Aware Locomotion Mode Recognition System for Powered Lower Limb Prostheses

Area of Science:

• Robotics
• Biomedical Engineering
• Human-Computer Interaction

Background:

• Powered artificial legs require sophisticated control systems for effective volitional movement.
• Locomotion mode (LM) recognition is crucial for adapting prosthetic function to different terrains.
• Existing systems often lack real-time environmental awareness, limiting adaptability.

Purpose of the Study:

• To develop and evaluate an environment-aware system for recognizing locomotion modes in powered artificial legs.
• To integrate terrain recognition (TR) with LM recognition for enhanced prosthetic control.
• To assess the system's performance in able-bodied individuals and a transfemoral amputee.

Main Methods:

• A portable terrain recognition (TR) module using an inertia measurement unit and laser distance meter was developed.
• A decision tree classifier was employed to identify terrain types and provide environmental information.
• Environmental data was integrated as a priori probability into a neuromuscular-mechanical-fusion-based LM recognition system.

Main Results:

• The TR module achieved over 98% accuracy and detected terrain transitions more than 500 ms in advance.
• Environmental information, coarse or refined, significantly improved LM recognition system performance.
• The environment-aware LM system demonstrated reliable performance even with noisy TR output, indicating robustness.

Conclusions:

• Environmental awareness is critical for the effective operation of wearable lower limb robotic devices.
• The developed system enables smoother locomotion mode transitions and enhances prosthetic control.
• This approach holds significant potential for improving prosthetic and orthotic device functionality in unconstructed environments.