ROS2SmolVLA: Enabling Small Vision-Language-Action Models for Integration into Industrial-Grade Lightweight Robots Paper • 2608.23320 • Published Aug 24 • 1