Event

CC-HGCBF Achieves Zero-Shot Safe Navigation for Heterogeneous Robots

1 reports · 1 sources

Joonkyung Kim et al. proposed the Capability-Conditioned Heterogeneous Graph Control Barrier Function (CC-HGCBF) and the Radial Capability Field (RCF) for safe navigation of heterogeneous mobile robot teams. The method establishes a unified safety representation for robots with different kinematic and actuation capabilities in a shared second-order reference-point operating space, using a fixed-dimensional characterization of each robot's state-dependent finite-horizon reachable velocity set, enabling the shared graph barrier and decentralized policy to adapt to changes in state, kinematic constraints, and control limits without predefined robot identities. The authors report that, after evaluation on mixed teams containing both seen and unseen robot models, experiments show the method achieves scalable safe coordination and zero-shot transfer to robot models and capability variations not present during training; physical robot experiments further show zero-shot deployment to unseen kinematic models without retraining.