Estimating Control Barriers from Offline Data

Estimating Control Barriers from Offline Data Hongzhan Y u 1, Seth Farrell 1, Ryo Y oshimitsu 2, Zhizhen Qin 1, Henrik I. Christensen 1 and Sicun Gao 1 Abstract -- Learning-based methods for constructing control barrier functions (CBFs) are gaining popularity for ensuring safe robot control. A major limitation of existing methods is their reliance on extensive sampling over the state space or online system interaction in simulation. In this work we propose a novel framework for learning neural CBFs through a fixed, sparsely-labeled dataset collected prior to training. Our approach introduces new annotation techniques based on out-of-distribution analysis, enabling efficient knowledge propagation from the limited labeled data to the unlabeled data. We also eliminate the dependency on a high-performance expert controller, and allow multiple sub-optimal policies or even manual control during data collection. We evaluate the proposed method on real-world platforms. With limited amount of offline data, it achieves state-of-the-art performance for dynamic obstacle avoidance, demonstrating statistically safer and less conservative maneuvers compared to existing methods. I NTRODUCTION Control Barrier Functions (CBFs) provide an effective framework for safe robot control [1], [2].The recent development of learning-based CBF methods exploit the expressiveness of neural networks and data-driven approaches to handle systems with complex dynamics and high uncertainty, with promising results [3], [4], [5], [6], [7], [8], [9]. However, the scalability of learning-based methods has been a major bottleneck. The typical approach for learning neural CBFs requires sampling over the entire state space to enforce constraints from the standard CBF conditions [10].