Co-Evolving Robot Orchestrators and Policies through Deployment

Hugging Face Daily Papers Papers
robotics vision-language-action vlm-orchestrator self-improvement fine-tuning real-world-deployment

Summary

提出了RoboCOP框架，让机器人编排器(VLM orchestrator)与视觉-语言-动作(VLA)策略在部署过程中协同进化，通过自采集演示数据在验证通过时微调策略，实现自改进飞轮，在仿真和真实世界任务上显著提升了泛化成功率。

Vision-language-action (VLA) policies trained on large datasets are capable within their training domains, yet they still fail to generalize to the variety of situations a robot meets in real-world deployment. Agentic robot systems complement the policy with a vision-language model (VLM) orchestrator that learns when to call the policy, how to instruct it, and when to use scripted skills instead. However, because the harness is built around a frozen policy that has limited language steerability, the orchestrator can avoid the policy's failures but never overcome them. The policy becomes the bottleneck of the whole system. Fine-tuning the policy can remove this bottleneck, but updating it alone decouples it from an orchestrator tuned to its old behavior. We propose Robo-COP, in which the orchestrator and policy co-evolve during deployment. Robo-COP curates skill demonstrations from its own executions, fine-tunes the policy when this data can address recurring failures, and adopts each new policy only after it improves the skills it was trained for. Across ten simulated RoboLab tasks, Robo-COP raises mean held-out success from 64.8% to 73.8% over the same harness with a frozen policy, while fine-tuning on a fixed schedule without verification reaches only 65.8%. On three real-world tasks, Robo-COP raises held-out success from 38.3% to 50.0%. Robo-COP turns deployment into a self-improving flywheel in which robots learn by doing, with each improvement in execution producing better data for the next round of learning. Videos and code are available at https://robo-cop.pages.dev/.
Original Article
View Cached Full Text

Cached at: 10/08/26, 12:15 PM

Paper page - Co-Evolving Robot Orchestrators and Policies through Deployment

Source: https://huggingface.co/papers/2610.09228

Abstract

Vision-language-action(VLA)policiestrainedonlargedatasetsarecapablewithintheirtrainingdomains,yettheystillfailtogeneralizetothevarietyofsituationsarobotmeetsinreal-worlddeployment.Agenticrobotsystemscomplementthepolicywithavision-languagemodel(VLM)orchestratorthatlearnswhentocallthepolicy,howtoinstructit,andwhentousescriptedskillsinstead.However,becausetheharnessisbuiltaroundafrozenpolicythathaslimitedlanguagesteerability,theorchestratorcanavoidthepolicy’sfailuresbutneverovercomethem.Thepolicybecomesthebottleneckofthewholesystem.Fine-tuningthepolicycanremovethisbottleneck,butupdatingitalonedecouplesitfromanorchestratortunedtoitsoldbehavior.WeproposeRobo-COP,inwhichtheorchestratorandpolicyco-evolveduringdeployment.Robo-COPcuratesskilldemonstrationsfromitsownexecutions,fine-tunesthepolicywhenthisdatacanaddressrecurringfailures,andadoptseachnewpolicyonlyafteritimprovestheskillsitwastrainedfor.AcrosstensimulatedRoboLabtasks,Robo-COPraisesmeanheld-outsuccessfrom64.8%to73.8%overthesameharnesswithafrozenpolicy,whilefine-tuningonafixedschedulewithoutverificationreachesonly65.8%.Onthreereal-worldtasks,Robo-COPraisesheld-outsuccessfrom38.3%to50.0%.Robo-COPturnsdeploymentintoaself-improvingflywheelinwhichrobotslearnbydoing,witheachimprovementinexecutionproducingbetterdataforthenextroundoflearning.Videosandcodeareavailableathttps://robo-cop.pages.dev/.

View arXiv pageView PDFAdd to collection

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2610.09228 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2610.09228 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2610.09228 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

CoDA: Co-Evolution of Data Engines and Action Policies

Hugging Face Daily Papers

CoDA 提出一种仿真数据引擎与机器人动作策略协同演进的框架，通过共享失败与学习进展记忆、视觉语言诊断以及 Text2potential 分阶段势能奖励，将机器人操作任务的分布外成功率从 5.3% 提升至 27.6%，并在 LIBERO-Pro 上取得显著改善。

Embodied Turing Machines: Stateful Code for Robot Recursive Self-Improvement

Hugging Face Daily Papers

本文提出 Code-Only-As-Policy (COAP) 范式，将机器人本体视为图灵机，用显式代码替代 VLM/VLA 进行决策与状态跟踪，并借助编码智能体实现闭环递归自我改进；在 RoboDojo 的 42 项双臂任务上达到 70.24% 成功率，且测试时无需模型在环。