Hierarchical Intention Tracking with Switching Trees for Real-Time Adaptation to Dynamic Human Intentions during Collaboration

2025-06-08

AI Summary

To address the challenges of multi-level, highly time-varying, and real-time human intent tracking in human-robot collaboration, this paper proposes a dynamic intent estimation framework integrating a variable-depth intent tree with Bayesian filtering. The method introduces a bidirectional probabilistic propagation mechanism—comprising upward measurement updates and downward posterior inference—to enable joint modeling and seamless switching among task-level, interaction-level, and verification-level intents. Hierarchical state propagation coupled with online correction ensures millisecond-scale intent tracking and proactive intent validation. Experimental evaluation on an assembly task demonstrates 100% task completion rate, a 23.6% improvement in operational efficiency, a 31.2% reduction in user physical workload, and a 47.8% decrease in interruption frequency. These results significantly enhance operator trust and comfort, outperforming state-of-the-art approaches in comprehensive performance.

Technology Category

• Intelligent Robots: State Estimation
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Activity and Plan Recognition

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

During collaborative tasks, human behavior is guided by multiple levels of intentions that evolve over time, such as task sequence preferences and interaction strategies. To adapt to these changing preferences and promptly correct any inaccurate estimations, collaborative robots must accurately track these dynamic human intentions in real time. We propose a Hierarchical Intention Tracking (HIT) algorithm for collaborative robots to track dynamic and hierarchical human intentions effectively in real time. HIT represents human intentions as intention trees with arbitrary depth, and probabilistically tracks human intentions by Bayesian filtering, upward measurement propagation, and downward posterior propagation across all levels. We develop a HIT-based robotic system that dynamically switches between Interaction-Task and Verification-Task trees for a collaborative assembly task, allowing the robot to effectively coordinate human intentions at three levels: task-level (subtask goal locations), interaction-level (mode of engagement with the robot), and verification-level (confirming or correcting intention recognition). Our user study shows that our HIT-based collaborative robot system surpasses existing collaborative robot solutions by achieving a balance between efficiency, physical workload, and user comfort while ensuring safety and task completion. Post-experiment surveys further reveal that the HIT-based system enhances the user trust and minimizes interruptions to user's task flow through its effective understanding of human intentions across multiple levels.

Problem

• Track dynamic hierarchical human intentions in real-time
• Adapt to evolving task and interaction preferences
• Balance efficiency, workload, and comfort in collaboration

Innovation

• Hierarchical Intention Tracking with Bayesian filtering
• Dynamic switching between Interaction-Task trees
• Multi-level intention coordination for user trust