Improving Walking Economy With an Ankle Exoskeleton Prior to Human-in-the-Loop Optimization.

Area of Science:

• Robotics
• Biomechanics
• Human-Computer Interaction

Background:

• Lower limb robotic exoskeletons aid locomotion and rehabilitation.
• Human-in-the-loop optimization customizes exoskeleton assistance but is time-consuming and can yield inaccurate results due to discomfort.
• Investigating pre-optimization assistance patterns can improve efficiency and user experience.

Purpose of the Study:

• To evaluate the effectiveness of different ankle exoskeleton assistance patterns on improving walking economy before human-in-the-loop optimization.
• To systematically assess biomechanical and physiological responses to various assistance patterns.
• To identify an optimal initial assistance pattern for exoskeleton optimization.

Main Methods:

• Experiments were conducted using a lightweight cable-driven ankle exoskeleton during walking.
• Nine assistance patterns were designed, varying peak torque magnitude and timing.
• Metabolic cost, soleus (SOL) muscle activity, joint kinematics, and joint moments were measured.

Main Results:

• One assistance pattern reduced metabolic cost by 17.1 ± 7.6% and soleus muscle activity by 40.9 ± 19.8%.
• Exoskeleton assistance altered ankle joint angles (dorsiflexion and plantarflexion) and reduced biological ankle moments.
• The most effective pattern featured 48% peak torque timing and 0.75 Nm/kg peak torque magnitude.

Conclusions:

• A specific ankle exoskeleton assistance pattern can significantly improve walking economy and reduce muscle effort.
• This identified pattern is suitable as an initial setting for human-in-the-loop optimization.
• Findings provide insights into human responses to exoskeleton assistance, guiding future optimization strategies.