A Haptic Shared-Control Architecture for Guided Multi-Target Robotic Grasping

Area of Science:

• Robotics
• Human-Computer Interaction
• Nuclear Engineering

Background:

• Robotic telemanipulation is crucial for the nuclear industry, particularly for decommissioning legacy radioactive waste.
• Current solutions, primarily mechanically linked master-slave manipulators, have seen limited advancement.
• There is a critical need for advanced robotic systems to accelerate waste management operations.

Purpose of the Study:

• To present a novel haptic shared-control approach for assisting human operators in object sorting and segregation tasks.
• To enhance the efficiency and effectiveness of remote handling in cluttered and unknown environments.
• To reduce the time and improve the precision of decommissioning tasks.

Main Methods:

• Utilized 3D scanning to identify potential grasp candidates on objects.
• Generated smooth and continuous guiding haptic cues for operator assistance.
• Implemented a shared-control system integrating human input with robotic guidance.
• Conducted human-subject studies with 15 participants to validate the approach.

Main Results:

• Demonstrated an average improvement of 20.8% in task completion time.
• Achieved a 20.1% reduction in linear trajectory, indicating increased precision.
• Reported a 32.5% increase in perceived effectiveness by the operators.
• Validated the system's ability to provide smooth transitions between tasks and objects.

Conclusions:

• The proposed haptic shared-control approach significantly outperforms standard teleoperation for nuclear waste handling tasks.
• This technology offers a promising solution for accelerating the decommissioning of legacy radioactive waste.
• The system enhances operator performance through intuitive haptic guidance, improving efficiency and safety.