A Mixed-Reality-Based Unknown Space Navigation Method of a Flexible Manipulator.

Area of Science:

• Robotics
• Human-Computer Interaction
• Computer Vision

Background:

• Hyper-redundant flexible manipulators possess high degrees of freedom (DoF) and adaptability, making them suitable for complex environments like debris rescue and pipeline inspection.
• Current manipulators lack the intelligence to autonomously navigate complex, unknown spaces, necessitating human intervention for control and decision-making.

Purpose of the Study:

• To design an interactive navigation method using mixed reality (MR) for hyper-redundant flexible manipulators operating in unknown environments.
• To develop a novel teleoperation system frame that enhances operator control and situational awareness.

Main Methods:

• Implemented a mixed reality (MR) interface displaying a virtual model of the remote workspace and interactive controls.
• Utilized a simultaneous localization and mapping (SLAM) algorithm with an RGB-D camera for environmental modeling.
• Employed an artificial potential field (APF) based method for path-finding and obstacle avoidance.

Main Results:

• The developed MR-based teleoperation system demonstrated good real-time performance.
• Simulations and experiments validated the system's accuracy, security, and user-friendliness.
• The system enables operators to observe remote situations and issue commands effectively, ensuring safe manipulator movement.

Conclusions:

• The proposed interactive navigation method effectively addresses the control challenges of hyper-redundant flexible manipulators in unknown spaces.
• The integration of MR, SLAM, and APF provides a robust and user-friendly solution for teleoperation.
• The system enhances safety and efficiency in complex robotic missions requiring human guidance.