LIBERO-Safety: A Comprehensive Benchmark for Physical and Semantic Safety in Vision-Language-Action Models
This paper introduces a safety benchmark and dataset to evaluate the safety of Vision-Language-Action models.
This paper introduces a new safety benchmark and dataset for evaluating the safety of Vision-Language-Action models, which is novel compared to prior work.
Before reading this…
Applications
- →Robotics
- →Autonomous systems
To understand this paper, make sure you know these concepts first:
- Robot learningfind papers →
- Vision-Language-Action modelsfind papers →
Abstract
More Like ThisDespite the impressive manipulation capabilities of Vision-Language-Action (VLA) models, their operational safety under strict constraints remains largely unverified. To address this, we introduce a parametric safety benchmark to procedurally generate safety-critical scenarios with comprehensive stochasticity. To overcome the scalability bottlenecks of human teleoperation, we develop a novel keypose-driven data generation pipeline. Leveraging this infrastructure, we curate a large-scale dataset of 19,664 strictly collision-free demonstrations with extensive domain randomization. We then conduct a systematic cross-paradigm evaluation of eight VLA and two embodied foundation models. Our analysis reveals a critical generalization-safety tension: although high-diversity training fosters safer trajectories, task success remains fundamentally bottlenecked by sub-optimal trajectory synthesis and semantic misalignment. By providing a scalable pipeline, a robust dataset, and profound failure-mode insights, LIBERO-Safety establishes a crucial foundation for developing safe and reliable VLA models.