Estimating Trust in Human-Robot Collaboration Through Behavioral Indicators and Explainability

2025-10-01

IEEE Robotics and Automation Letters

AI Summary

This work proposes a data-driven framework for trust evaluation and optimization to enhance safety and comfort in human–robot collaboration within Industry 5.0. By integrating operator behavioral metrics with preference-based feedback, the authors develop an interpretable trust prediction model and employ a preference optimization algorithm to dynamically generate robot trajectories that maximize human trust. The key innovation lies in the novel integration of behavioral indicators with interpretable modeling for trust assessment, coupled with human-preference-driven trajectory optimization. Experimental validation on a chemical mixing task demonstrates that the proposed voting classifier achieves a trust prediction accuracy of 84.07% (AUC-ROC = 0.90), significantly outperforming baseline methods.

Technology Category

• Humans and AI: Learning Human Values and Preferences
• Intelligent Robots: Human-Robot Interaction
• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness

Application Category

• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success

Abstract

Industry 5.0 focuses on human-centric collaboration between humans and robots, prioritizing safety, comfort, and trust. This study introduces a data-driven framework to assess trust using behavioral indicators. The framework employs a Preference-Based Optimization algorithm to generate trust-enhancing trajectories based on operator feedback. This feedback serves as ground truth for training machine learning models to predict trust levels from behavioral indicators. The framework was tested in a chemical industry scenario where a robot assisted a human operator in mixing chemicals. Machine learning models classified trust with over 80% accuracy, with the Voting Classifier achieving 84.07% accuracy and an AUC-ROC score of 0.90. These findings underscore the effectiveness of data-driven methods in assessing trust within human-robot collaboration, emphasizing the valuable role behavioral indicators play in predicting the dynamics of human trust.

Problem

• trust estimation
• human-robot collaboration
• behavioral indicators
• Industry 5.0
• human trust

Innovation

• Trust Estimation
• Behavioral Indicators
• Preference-Based Optimization
• Human-Robot Collaboration
• Explainable AI