Hybrid Analytical and Simulation-Based Approach for Workspace Verification of a Pneumatic Upper Limb Exoskeleton

Area of Science:

• Robotics and Mechanical Engineering
• Human-Computer Interaction
• Biomechanics

Background:

• Designing active pneumatic upper limb exoskeletons presents challenges in defining a safe workspace.
• Existing analytical kinematic methods fail to detect geometric collisions in closed kinematic chains, risking mechanical damage and user safety.

Purpose of the Study:

• To propose and evaluate a hybrid algorithm for verifying the workspace of pneumatic exoskeletons.
• To ensure kinematic safety and prevent collisions within the exoskeleton's operational space.

Main Methods:

• A hybrid algorithm combining analytical modeling (Product of Exponentials method in MATLAB) and high-performance static simulation (Unity environment).
• Generation of 758 million discrete manipulator positions, followed by a two-stage filtering process: analytical verification and physical collision detection using the PhysX engine.

Main Results:

• The analytical model filtered 99.6% of invalid configurations, but up to 50% of remaining positions caused critical geometric collisions.
• The hybrid algorithm achieved full static workspace verification in under 20 minutes.
• Established a reachable zone topology with asymmetry and a "manoeuvrability core" in the anterior hemisphere.

Conclusions:

• Analytical kinematics must be complemented by physical collision detection for designing hybrid kinematic mechanisms.
• The developed algorithm generates verified, kinematically safe exoskeleton states, forming a basis for hierarchical control systems.
• The approach is applicable for verifying collision-free trajectories in robotic systems simulations.