Related Experiment Video
Updated: Jun 27, 2026
07:30
The Muscle Cuff Regenerative Peripheral Nerve Interface for the Amplification of Intact Peripheral Nerve Signals
Published on: January 13, 2022
Multimodal EEG-EMG and FEM-Based Adaptive Control of Passive Upper-Limb Exoskeletons
Luigi Bibbò1, Filippo Laganà2, Salvatore A Pullano2
1Department of Civil, Energy, Environment and Materials (DICEAM), "Mediterranea" University of Reggio Calabria, I-89124 Reggio Calabria, Italy.
Sensors (Basel, Switzerland)
|June 26, 2026
Summary
This study introduces a brain-machine interface for passive exoskeletons, combining EEG and EMG signals for adaptive assistance. It accurately interprets user effort and motor intent, enhancing wearable robotic functionality.
Area of Science:
- Robotics
- Neuroscience
- Biomechanical Engineering
Background:
- Wearable robotics require adaptive assistance for real-world tasks.
- Passive exoskeletons offer energy efficiency but lack inherent adaptability.
- Integrating neural and muscular signals can enable adaptive control.
Purpose of the Study:
- To propose a multimodal neural interface for passive upper-limb exoskeletons.
- To classify motor gestures and estimate cognitive/muscular effort using EEG and EMG signals.
- To achieve software-level adaptive assistance without active actuation.
Main Methods:
- Implemented a hybrid CNN-LSTM deep fusion architecture.
- Used synchronous EEG-EMG acquisition via the LiveAmp platform.
- Employed Finite Element Method (FEM) for biomechanical modeling and analysis.
Main Results:
- Achieved 90% average classification accuracy and 0.85 F1-score.
- Demonstrated inference latency below 180 ms for real-time applicability.
- Validated adaptive assistance modulation via cognitive indices (CLI, FAI).
Conclusions:
- The multimodal interface enables effective software-level adaptation in passive exoskeletons.
- Findings advance brain-machine interfaces for energy-efficient, adaptive wearable robotics.
- The approach shows potential for rehabilitation, occupational ergonomics, and human-robot interaction.
