Human-robot mutual adaptation in collaborative tasks: Models and experiments.

Area of Science:

• Robotics
• Human-Computer Interaction
• Artificial Intelligence

Background:

• Effective team collaboration is essential for complex tasks.
• Existing human-robot interaction models often lack mutual adaptation.
• Understanding human adaptive behaviors is key to improving team performance.

Purpose of the Study:

• To propose a computational formalism for mutual adaptation in human-robot teams.
• To develop a model capturing human adaptive behaviors under bounded memory.
• To enhance the effectiveness and trust in human-robot collaborative tasks.

Main Methods:

• Introduced the Bounded-Memory Adaptation Model (BMAM), a probabilistic finite-state controller.
• Integrated the BMAM into a probabilistic decision process for robot guidance.
• Conducted human subject experiments to evaluate the proposed formalism.

Main Results:

• The proposed formalism significantly improved human-robot team effectiveness compared to one-way adaptation.
• Mutual adaptation, facilitated by the BMAM, led to better task completion outcomes.
• Human participants maintained trust in the robot when mutual adaptation was employed.

Conclusions:

• Mutual adaptation is crucial for optimizing human-robot team performance.
• The Bounded-Memory Adaptation Model provides a viable computational approach for mutual adaptation.
• This work advances the field of human-robot collaboration by enabling more effective and trustworthy teams.