Informed circular fields: a global reactive obstacle avoidance framework for robotic manipulators.

Area of Science:

• Robotics
• Artificial Intelligence
• Control Systems

Background:

• Robotic manipulators require advanced motion planning for safe navigation in complex, dynamic environments.
• Existing methods often struggle with real-time adaptation and comprehensive obstacle avoidance.
• Integrating global and local planning strategies is crucial for robust performance.

Purpose of the Study:

• To develop a unified framework for global reactive motion planning in robotic manipulators.
• To enhance obstacle avoidance capabilities by combining global trajectory insights with local reactive control.
• To improve the adaptability and safety of robots operating in unpredictable, human-populated workspaces.

Main Methods:

• Utilizing local minima-free circular fields for reactive control command generation.
• Leveraging global environmental information from configuration space motion planners to guide trajectories.
• Extending the virtual agents framework to simulate multiple robot trajectories for enhanced avoidance.
• Integrating vision feedback for real-world robot control and adaptation.

Main Results:

• The proposed framework consistently outperforms existing motion planners in over 4,000 simulation scenarios.
• Demonstrated effective and prompt adaptation to unpredictable human movements in real-world experiments with a Franka Emika robot.
• Achieved comprehensive obstacle avoidance for the entire robotic manipulator body.

Conclusions:

• The developed global reactive motion planning framework offers a robust and versatile solution for dynamic environments.
• The unified approach effectively combines global planning with local reactive control for superior performance.
• Validated efficacy through extensive simulations and real-world robotic experiments.