Benchmarking the Effects on Human-Exoskeleton Interaction of Trajectory, Admittance and EMG-Triggered Exoskeleton...

Area of Science:

• Biomedical Engineering
• Rehabilitation Robotics
• Human-Robot Interaction

Background:

• Robotic gait training is increasingly used but its effectiveness and impact on user experience are debated.
• Current control strategies often fail to integrate user intention, limiting rehabilitation potential.
• Understanding how different exoskeleton controllers affect muscle coordination, effort, and acceptance is crucial.

Purpose of the Study:

• To benchmark three exoskeleton control strategies: trajectory assistance (TC), compliant assistance (AC), and EMG-Onset stepping control (OC).
• To evaluate the impact of these controllers on kinematics, muscle activation, physiological responses, and user-perceived effort.
• To assess the potential of novel EMG-based control for gait rehabilitation.

Main Methods:

• Seven healthy volunteers participated in an exploratory study using the EUROBENCH facility.
• Exoskeleton kinematics, electromyography (EMG) for muscle activation, heart rate, breathing rate, and skin conductance were recorded.
• User-perceived effort was quantified, and data analysis followed EUROBENCH protocols.

Main Results:

• The OC controller robustly detected stepping intention, even with noisy EMG data.
• AC and OC controllers induced similar kinematic changes compared to TC.
• Muscle synergies showed minor alterations, increased agonist-antagonist co-contraction, and reduced activation duration with OC, without affecting overall physiological cost or user perception.

Conclusions:

• While AC and OC controllers allow users to modulate their gait, they did not significantly change the physiological cost or subjective experience.
• Both AC and OC controllers show promise as gait rehabilitation tools.
• This study provides valuable benchmarking data for human-exoskeleton interaction and introduces a novel EMG-based controller.