UMBRELLA: Uncertainty-aware Multi-roBot REactive Coordination under Dynamic TemporaL Logic TAsk

Abstract

This paper presents UMBRELLA, a novel framework for uncertainty-aware multi-robot reactive coordination under dynamic temporal logic tasks. The proposed approach addresses the challenge of coordinating multiple robots in dynamic environments where tasks are specified using temporal logic and uncertainty must be explicitly considered. Our framework combines reactive planning with uncertainty quantification to enable robust multi-robot coordination.

Key Contributions

• Uncertainty-aware Coordination: A novel framework that explicitly considers uncertainty in multi-robot coordination
• Dynamic Temporal Logic Tasks: Support for temporal logic task specifications that can change dynamically
• Reactive Planning: Real-time adaptation to environmental changes and task modifications
• Comprehensive Evaluation: Validation through both simulation and hardware experiments

Experimental Results

Our evaluation demonstrates the effectiveness of UMBRELLA across three scenarios:

1. Task Planning: 12 robots coordinating to track 4 dynamic targets across 12 tasks in multiple scenes
2. ROS Simulation: 8 robots and 3 dynamic targets executing 10 tasks in a city-like environment
3. Hardware Experiments: 4 physical robots and 2 dynamic targets performing 7 tasks with real-world constraints

The results show significant improvements in task completion rates and coordination efficiency compared to baseline methods.

Code and Data

The implementation and experimental data will be available soon.

Recommended citation: Zhao, Q., Guo, M., Du, H., Lindemann, L., & Li, Z. (2026). UMBRELLA: Uncertainty-aware Multi-roBot REactive Coordination under Dynamic TemporaL Logic TAsk. IEEE International Conference on Robotics and Automation (ICRA).
Download Paper