Coordinating With a Robot Partner Affects Neural Processing Related to Action Monitoring

Area of Science:

• Robotics
• Neuroscience
• Human-Computer Interaction

Background:

• Robots are increasingly integrated into social environments, necessitating an understanding of human-robot interaction dynamics.
• Investigating the behavioral and neural differences between human-human and human-robot collaboration is crucial for seamless integration.

Purpose of the Study:

• To compare human-human and human-robot interactions behaviorally and neurophysiologically.
• To identify neural underpinnings of human responses during collaborative tasks with robots.

Main Methods:

• Adapted a collaborative dynamical paradigm involving 12 participants guiding a tablet with another human or a robot.
• Introduced perturbations requiring corrective actions and recorded electroencephalography (EEG) data.
• Analyzed behavioral data (velocity, positional error) and event-related potentials from EEG.

Main Results:

• Human-human interactions showed increased velocity and positional errors compared to human-robot interactions.
• Electroencephalography revealed significant differences in neural processing between partners, particularly at fronto-central electrodes.
• Event-related potential amplitudes were stronger with robot partners, indicating distinct neural activity.

Conclusions:

• Coordinating with robots influences action monitoring and error processing differently than interacting with humans.
• Human participants exhibit unique responses to errors in human-robot interactions.
• Real-time neural activity analysis can enhance human-robot communication.