Decision-Making Model for Adaptive Impedance Control of Teleoperation Systems

Area of Science:

• Robotics
• Human-Computer Interaction
• Cognitive Science

Background:

• Teleoperation systems require effective human-machine interaction, especially in partially structured environments.
• Current haptic assistance strategies often lack task and situation-specific adaptability.
• Decision-making models from cognitive science offer potential for enhanced robotic control but are underexplored in assisted robotics.

Purpose of the Study:

• To propose and evaluate a novel haptic assistance strategy for teleoperation.
• To integrate decision-making models with impedance control for adaptive haptic shared control.
• To investigate the effectiveness of the Drift-Diffusion model for scheduling haptic assistance.

Main Methods:

• Developed a haptic assistance strategy using an admittance controller scheduled by a decision-making model.
• Combined the decision-making model with standard impedance control techniques for bilateral teleoperation.
• Utilized the Drift-Diffusion model as the core scheduling strategy for haptic shared control.
• Defined and manipulated reward functions to adapt the assistance mechanism.

Main Results:

• Experimental data validated the Drift-Diffusion model's suitability for scheduling haptic shared control.
• The proposed strategy demonstrated task and situation-specific compromises between tracking performance and human-machine interaction.
• User study results showed the influence of reward structure on realized haptic assistances.

Conclusions:

• The Drift-Diffusion model provides an effective framework for adaptive haptic shared control in teleoperation.
• The developed strategy enhances human-machine interaction and performance in partially structured environments.
• Guidelines for tuning the decision-making model and reward functions are presented, offering practical insights for future research and application.