WBAG: A Whole-Body and Attached-Geometry Safety Framework for Vision-Language-Action Manipulation

2026-10-01

AI Summary

This study addresses the collision risks arising from unmodeled whole-body and grasped-object geometries during Vision-Language-Action (VLA) policy deployment by proposing a safety framework based on Control Barrier Functions (CBFs). The method explicitly models the robot's full-body and attached geometries to construct grasp-conditioned safe sets, transforming dynamic geometric representations into differentiable CBF constraints. By minimizing action modifications, it achieves adaptive obstacle avoidance that responds dynamically to changing grasp states. Evaluated on the SafeLIBERO benchmark, the proposed approach attains a 97.38% scene safety rate and a 59.38% safe success rate, significantly outperforming existing baselines.

Abstract

Vision-language-action (VLA) policies have demonstrated impressive capabilities in generalizable robotic manipulation, but their deployment in the real world remains challenging due to potential collisions involving different parts of the robot, manipulated objects, and the surrounding environment. Existing inference-time VLA safety frameworks typically rely on simplified end-effector-centered representations that do not explicitly model the full articulated robot and attached-object geometry. In this paper, we present WBAG, a safety framework that models the robot's whole-body and grasp-dependent attached geometry. WBAG constructs a grasp-conditioned safe set that adapts the protected geometry as objects are grasped, then converts this evolving geometry into differentiable CBF constraints that minimally modify the VLA's native six-dimensional operational-space action for collision avoidance across robot, scene, and attached geometry. On the SafeLIBERO benchmark, a variant of LIBERO augmented with obstacles for safety evaluation, WBAG achieves the best overall safety and safe task success among the evaluated methods under a scene-level safety evaluator that monitors all eligible non-task objects, reaching 97.38\% aggregate Scene Safety and 59.38\% Safe Success.

Problem

• Vision-Language-Action
• Robotic Manipulation
• Collision Avoidance
• Safety Framework
• Whole-Body Geometry

Innovation

• Vision-Language-Action
• Whole-Body Geometry
• Control Barrier Functions
• Robotic Manipulation Safety
• Differentiable Constraints