发表机构
Nanyang Technological University; University of Washington; University of California, Berkeley; KAIST（南洋理工大学; 华盛顿大学; 加州大学伯克利分校; 韩国科学技术院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出SafeStreamingFlow，通过对齐采样与执行动力学实现安全流式规划，降低延迟并提升安全性，同时保持目标到达成功率。
AI 中文摘要
基于扩散/流匹配的生成式规划器可以从示范中学习合成长期轨迹。然而，实际部署需要（i）在执行过程中强制执行安全约束，以及（ii）在快速执行速率下进行紧密的在线重规划。先前的安全扩散/流规划器一次性生成智能体的完整轨迹，同时反复扰动中间状态以满足安全约束。这种方法不仅计算密集，而且由于学习到的采样动力学与系统的执行动力学不同，引入了分布偏移。我们提出SafeStreamingFlow，一种目标条件规划器，通过顺序集成学习到的状态向量场与分层状态预测，将流采样动力学与执行动力学对齐。重要的是，我们只需通过高阶控制屏障函数对执行步骤强制执行安全约束。在导航、竞速和运动基准测试中，SafeStreamingFlow相比现有方法降低了规划延迟并提高了安全性，同时保持了有竞争力的目标到达成功率。
英文摘要
Generative planners based on diffusion/flow matching can learn to synthesize long-horizon trajectories from demonstrations. However, real-world deployment requires (i) enforcing safety constraints during execution and (ii) tight online replanning at fast execution rates. Prior safe diffusion/flow planners generate the agent's full trajectory at once, while repeatedly perturbing intermediate states to satisfy safety constraints. This approach is not only computationally intensive, but also introduces distribution shift since the learned sampling dynamics is distinct from the system's execution dynamics. We propose SafeStreamingFlow, a goal-conditioned planner that aligns flow sampling dynamics with execution dynamics by sequentially integrating a learned state vector field with hierarchical state prediction. Importantly, we need to enforce safety constraints only for the executed step via high order control barrier functions. Across navigation, racing, and locomotion benchmarks, SafeStreamingFlow reduces planning latency and improves safety compared to existing methods, while maintaining competitive goal-reaching success.
CommentsAccepted to the 10th Conference on Robot Learning (CoRL 2026). Project page: https://jang-seunghwan.github.io/SafeStreamingFlowPlanning/
