AI & ML New Capability

A flow-based navigation policy that achieves zero-shot sim-to-real transfer across wheeled, quadrupedal, and humanoid platforms.

arXiv · March 16, 2026 · 2603.12806

Zeying Gong, Yangyi Zhong, Yiyi Ding, Tianshuai Hu, Guoyang Zhao, Lingdong Kong, Rong Li, Jiadi You, Junwei Liang

Why it matters

FLUX linearizes probability flow to replace iterative denoising, making inference 47% faster than prior methods. Its ability to generalize across radically different robotic embodiments without fine-tuning represents a major step toward a foundation model for autonomous navigation.

From the abstract

Autonomous navigation requires a broad spectrum of skills, from static goal-reaching to dynamic social traversal, yet evaluation remains fragmented across disparate protocols. We introduce DynBench, a dynamic navigation benchmark featuring physically valid crowd simulation. Combined with existing static protocols, it supports comprehensive evaluation across six fundamental navigation tasks. Within this framework, we propose FLUX, the first flow-based unified navigation policy. By linearizing pro