Digital Guinea Pig: Merits and Methods of Human-in-the-Loop Simulation for Upper-Limb Exoskeletons
IEEE ... International Conference on Rehabilitation Robotics : [Proceedings]
|September 30, 2022
Summary
This study introduces a physics simulation for human-exoskeleton interaction, enhancing hardware and software development. The simulation accurately predicts system responses, improving safety and efficiency in exoskeleton design.
Area of Science:
- Robotics
- Human-Computer Interaction
- Biomechanics
Background:
- Exoskeletons require careful design due to continuous haptic interaction with human limbs.
- Validating exoskeleton hardware and control policies often involves complex and potentially hazardous real-world experiments.
Purpose of the Study:
- To present a physics simulation framework for validating the software stack of human-exoskeleton systems.
- To model and predict the closed-loop haptic interaction dynamics between a human and a 9-DOF exoskeleton arm.
Main Methods:
- Developed a rigid-body dynamics simulation of a coupled human-exoskeleton arm.
- Modeled human-robot interaction using decoupled spring-damper systems based on anthropometric data.
- Validated the simulation by predicting the system's response to impacts and assessing haptic-rendering performance.
Main Results:
- The simulation accurately predicted intrusion into compliant walls (6-13% relative accuracy).
- Predicted admissible control gains with approximately 14% accuracy.
- Demonstrated the simulation's capability to model closed-loop system reactions to external stimuli.
Conclusions:
- The simulation framework is valuable for validating exoskeleton prototype software and control policies.
- It aids in developing intuition and understanding complex coupled human-robot dynamics.
- While quantitative predictions have limitations, the simulation enhances development efficiency and safety.
