Related Experiment Video
Updated: Jun 12, 2025

Characterization of the Sense of Agency over the Actions of Neural-machine Interface-operated Prostheses
Published on: January 7, 2019
Self-assessment in machines boosts human Trust.
Dana Warmsley1, Krishna Choudhary1, Jocelyn Rego1
1Intelligent Systems Center, HRL Laboratories, Malibu, CA, United States.
Machines can improve human trust and team performance by assessing their own capabilities in real-time. This trust calibration system enhances human-machine collaboration, boosting adoption of autonomous systems.
Area of Science:
- Human-Computer Interaction
- Artificial Intelligence
- Robotics
Background:
- Low trust in autonomous systems hinders their widespread adoption and effectiveness.
- Current trust calibration methods often neglect the machine's self-assessment capabilities.
- Effective human-machine collaboration requires synchronized trust levels.
Purpose of the Study:
- To develop and evaluate a closed-loop trust calibration system for human-machine collaboration.
- To investigate the impact of machine self-assessment on human trust and team performance.
- To demonstrate the system's applicability in semi-autonomous image classification tasks.
Main Methods:
- Implemented a closed-loop system where machines assess their capabilities and human trust in real-time.
- Designed a human-machine collaboration task for image classification.
- Compared a trained machine self-assessment approach against a baseline without it.
Main Results:
- Achieved approximately 40% improvement in human trust levels.
- Observed a 5% increase in overall team performance.
- Demonstrated these gains were achieved with identical machine performance levels between conditions.
Conclusions:
- Machine self-assessment is a critical component for effective trust calibration in human-machine systems.
- The developed trust calibration system significantly enhances both trust and performance in collaborative tasks.
- The system is adaptable to various semi-autonomous applications requiring human-machine interaction.
Related Concept Videos
Self-Evaluation: Self-Enhancement and Self-Verification
The Sense of Self: Reflected Self-Appraisal and Social Comparison
Self-Schemas
Stereotype Content Model
Self-Presentation: Self-Monitoring and Self-Handicapping
Trait and State Self-Esteem

