Adaptive robot climbing with magnetic feet in unknown slippery structure

Area of Science:

• Robotics
• Control Systems
• Artificial Intelligence

Background:

• Maintaining firm foot contact is critical for climbing robots, especially in hazardous high-altitude operations.
• Reduced adhesion due to environmental factors poses significant risks, potentially leading to catastrophic failures.

Purpose of the Study:

• To develop a robust planning and control framework for climbing robots that ensures stability and safety in unknown environments, specifically addressing slippage.
• To enhance the real-time adaptability of climbing robots to uncertain contact conditions and dynamic environmental changes.

Main Methods:

• Center of mass (CoM) trajectory optimization under estimated contact conditions.
• Kalman filter-based approach for real-time estimation of uncertain environmental parameters and subsequent CoM trajectory re-planning.
• Online weight adaptation within a whole-body control (WBC) framework to adjust ground reaction force (GRF) distribution.

Main Results:

• The proposed CoM trajectory optimization achieved state-of-the-art fast computation through trajectory parameterization and linear algebra techniques.
• The framework demonstrated robustness to slippage by enabling real-time motion re-planning and GRF redistribution.
• Simulations using the magnetic-legged climbing robot Manegto validated the effectiveness of the proposed algorithm.

Conclusions:

• The developed framework significantly improves the safety and reliability of climbing robots by effectively managing slippage and adapting to unknown environments.
• Real-time parameter estimation and adaptive control strategies are crucial for robust robotic locomotion in challenging conditions.