CREPES-X: Robust Multi-Robot Relative Pose Estimation
Published:Dec 31, 2025 07:47
•1 min read
•ArXiv
Analysis
This paper presents CREPES-X, a novel system for relative pose estimation in multi-robot systems. It addresses the limitations of existing approaches by integrating bearing, distance, and inertial measurements in a hierarchical framework. The system's key strengths lie in its robustness to outliers, efficiency, and accuracy, particularly in challenging environments. The use of a closed-form solution for single-frame estimation and IMU pre-integration for multi-frame estimation are notable contributions. The paper's focus on practical hardware design and real-world validation further enhances its significance.
Key Takeaways
- •CREPES-X is a hierarchical relative localization framework for multi-robot systems.
- •It integrates bearing, distance, and inertial measurements.
- •The system is robust to outliers and performs well in challenging environments.
- •It uses a two-stage hierarchical estimator for speed, accuracy, and robustness.
- •Real-world experiments validate its effectiveness with low RMSE.
Reference
“CREPES-X achieves RMSE of 0.073m and 1.817° in real-world datasets, demonstrating robustness to up to 90% bearing outliers.”