arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.19983cs.ROcs.AI
收藏

面向开放世界机器人的上下文安全推理与 grounding

Contextual Safety Reasoning and Grounding for Open-World Robots

University of Pennsylvania(宾夕法尼亚大学);Carnegie Mellon University(卡内基梅隆大学)

Zachary Ravichandran, David Snyder, Alexander Robey, Hamed Hassani, Vijay Kumar, George J. Pappas

AI总结:

CORE框架通过视觉语言模型实现在线上下文推理与空间grounding,提供概率安全保证,在开放世界中有效执行上下文适应的安全行为。

摘要:

机器人正越来越多地在开放世界环境中运行,其中安全行为依赖于上下文:同一个走廊在拥挤与空旷、紧急情况与正常操作时可能需要不同的导航策略。传统安全方法强制在用户指定的上下文中使用固定约束,限制了其处理现实部署中开放性上下文变化的能力。我们通过CORE,一个安全框架,解决这一缺口,该框架能够在线进行上下文推理、grounding和执行,而无需事先了解环境(例如地图或安全规范)。CORE使用视觉语言模型(VLM)从视觉观察中持续推理上下文相关的安全规则,将这些规则 grounding 在物理环境中,并通过控制屏障函数执行由此产生的空间定义的安全集。我们为CORE提供了概率安全保证,考虑了感知不确定性,并通过模拟和现实世界实验证明,CORE在未见过的环境中执行上下文适当的行为主导,显著优于缺乏在线上下文推理的先前语义安全方法。消融研究验证了我们的理论保证,并突显了基于VLM的推理和空间grounding对于在新设置中执行上下文安全的重要性。我们还提供了额外资源:https://zacravichandran.github.io/CORE。

英文摘要:

Robots are increasingly operating in open-world environments where safe behavior depends on context: the same hallway may require different navigation strategies when crowded versus empty, or during an emergency versus normal operations. Traditional safety approaches enforce fixed constraints in user-specified contexts, limiting their ability to handle the open-ended contextual variability of real-world deployment. We address this gap via CORE, a safety framework that enables online contextual reasoning, grounding, and enforcement without prior knowledge of the environment (e.g., maps or safety specifications). CORE uses a vision-language model (VLM) to continuously reason about context-dependent safety rules directly from visual observations, grounds these rules in the physical environment, and enforces the resulting spatially-defined safe sets via control barrier functions. We provide probabilistic safety guarantees for CORE that account for perceptual uncertainty, and we demonstrate through simulation and real-world experiments that CORE enforces contextually appropriate behavior in unseen environments, significantly outperforming prior semantic safety methods that lack online contextual reasoning. Ablation studies validate our theoretical guarantees and underscore the importance of both VLM-based reasoning and spatial grounding for enforcing contextual safety in novel settings. We provide additional resources at https://zacravichandran.github.io/CORE.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑