ROS2SmolVLA: Enabling Small Vision-Language-Action Models for Integration into Industrial-Grade Lightweight Robots
2026-08-24 • Robotics
Robotics
AI summaryⓘ
The authors explain that as factories need to make more varied products in smaller amounts, robot systems must become more flexible. They adapted a small vision-language-action model called SmolVLA to work with real industrial robots, specifically Universal Robots UR10e, using ROS 2 software. They created an open-source tool to help connect this model to industrial robot hardware, making it easier to use both in labs and factories. Their tests show SmolVLA works well for simple robot tasks that need to be run locally on-site. This approach avoids issues seen with big models that require cloud computing or expensive lab robots.
Vision-Language-Action ModelsRobot AutomationUniversal Robots UR10eROS 2SmolVLAPick-and-Place TaskIndustrial RobotsOn-Premise ComputingOpen-Source SoftwareAdaptive Production Systems
Authors
Nils Mandischer, Noah Böckmann, Ludwig Holl, Lars Mikelsons
Abstract
Industrial demand changes the paradigms of production. Due to smaller batch sizes and more variations in products, companies face a growing challenge to adopt more adaptive production systems. In particular, robot-based automation is usually static and fails to respond to constantly changing processes. Vision-Language-Action (VLA) Models are a promising opportunity to mitigate this challenge by generating robot actions based on the observed system state. However, current research either focuses on large models that cannot be computed on premise, creating compliance and security challenges, or use lab-grade robot hardware that obscures exploitation in real industrial settings. In this work, we adapt Hugging Face's SmolVLA for Universal Robots lightweight robots. Further, we release the open-source repository ROS2SmolVLA that implements an interface for ROS 2 to SmolVLA, and makes it applicable for industrial-grade hardware. By this, we allow a lenient adoption into lab and industrial environments. We validate the functionality of SmolVLA for a Universal Robots UR10e using a pick-and-place task and give implementation guidelines. Our findings support that SmolVLA is a well-suited option for small-sized tasks that need to be computed on premise.