Towards using musculoskeletal models for intelligent control of physically assistive robots

Area of Science:

• Robotics
• Biomechanics
• Human-Computer Interaction

Background:

• Growing need for intelligent robotic systems in human assistance (e.g., rehabilitation, assistive living).
• Current control systems lack personalization and deep understanding of user's physical state.

Purpose of the Study:

• To propose and develop a musculoskeletal model for estimating human user strength.
• To utilize estimated strength for improving robot control schemes in physical human-robot interaction.
• To identify muscle-based weaknesses in task-specific scenarios.

Main Methods:

• Development of an optimization model based on a musculoskeletal model.
• Estimation of human strength in specific dynamic states.
• Analysis of muscle-based weaknesses in task space.

Main Results:

• Successful estimation of human strength using the developed musculoskeletal model.
• Identification of methods to observe task-specific muscle weaknesses.
• Discussion of integration challenges and potential solutions for robot control.

Conclusions:

• Musculoskeletal modeling offers a viable method for estimating human strength to personalize robot assistance.
• This approach can lead to more intelligent and effective robot control for rehabilitation and assistive living.
• Further research is needed to fully integrate these models into real-world robotic systems.