analiza aiqbench
SimVLA demonstrates a significant breakthrough in robotics by proving that Visual-Language-Action models can achieve zero-shot transfer to real-world mobile manipulation tasks using exclusively synthetic data. This approach effectively bypasses the high cost and complexity of teleoperation, showing that simulation-derived supervision can outperform policies trained on substantial amounts of real-world demonstrations. The work highlights the viability of scaling robotic learning through advanced simulation environments rather than physical trial-and-error.
SimVLA: Zero-Shot Sim-to-Real VLA Learning for Mobile Manipulation
arXiv:2610.11248v1 Announce Type: new Abstract: Large-scale, diverse datasets have driven the success of LLMs and VLMs. But VLAs for robotics remain limited by the cost and complexity of real-world data collection. While simulation offers a scalable alternative, its potential for sim-to-real VLA learning in mobile manipulation remains largely underexplored. We introduce SimVLA, an end-to-end framework that trains VLAs entirely on synthetic simulation data without teleoperation for mobile manipulation. SimVLA is first pre-trained on two complementary simulation-derived datasets: SimAction, a large-scale robot action dataset spanning 35 diverse mobile manipulation tasks, generated by composing atomic skills, and SimVQA, which leverages privileged simulator state to provide spatial, geometric, and subtask-level visual-language supervision. We further post-train SimVLA on a mixture of SimAction and SimDeploy, a dataset collected from policy rollouts across diverse simulated environments. We evaluate SimVLA on tasks including restocking, pouring, and cleaning, and show zero-shot transfer to real-world mobile manipulation, including real home environments. SimVLA outperforms policies trained on 50 in-domain real-world demonstrations, suggesting that simulation can enable scalable sim-to-real mobile manipulation. We further demonstrate the value of multiple complementary forms of supervision for effectively leveraging simulation in VLA training.
Abstract:Large-scale, diverse datasets have driven the success of LLMs and VLMs. But VLAs for robotics remain limited by the cost and complexity of real-world data collection. While simulation offers a scalable alternative, its potential for sim-to-real VLA learning in mobile manipulation remains largely underexplored. We introduce SimVLA, an end-to-end framework that trains VLAs entirely on synthetic simulation data without teleoperation for mobile manipulation. SimVLA is first pre-trained on two complementary simulation-derived datasets: SimAction, a large-scale robot action dataset spanning 35 diverse mobile manipulation tasks, generated by composing atomic skills, and SimVQA, which leverages privileged simulator state to provide spatial, geometric, and subtask-level visual-language supervision. We further post-train SimVLA on a mixture of SimAction and SimDeploy, a dataset collected from policy rollouts across diverse simulated environments. We evaluate SimVLA on tasks including restocking, pouring, and cleaning, and show zero-shot transfer to real-world mobile manipulation, including real home environments. SimVLA outperforms policies trained on 50 in-domain real-world demonstrations, suggesting that simulation can enable scalable sim-to-real mobile manipulation. We further demonstrate the value of multiple complementary forms of supervision for effectively leveraging simulation in VLA training.
Robotics (cs.RO)
arXiv:2610.11248 [cs.RO]
(or arXiv:2610.11248v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2610.11248 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 276 cuvinte
Subiecte conexe
Worlds in One Demo: A Synthetic Data Engine for Learning Open-World Mobile Manipulation
arXiv · 20 iul. 2026
RoboRender: Robot-Oriented Video Generation for Visual Sim-to-Real Transfer
arXiv · 8 oct. 2026
Exploratory, Communicative, and Deployable: Vision-Driven Embodied Agents for Open-World Mobile Manipulation
arXiv · 28 iul. 2026
SemAnCorr: Semantic Anchored Correspondence for Zero-Shot Manipulation Skill Transfer
arXiv · 31 iul. 2026