Maestro: Orchestrating Robotics Modules with Vision-Language Models for Zero-Shot Generalist Robots
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917089956069376 |
|---|---|
| author | Shi, Junyao Yang, Rujia Chao, Kaitian Wan, Selina Bingqing Shao, Yifei Lei, Jiahui Qian, Jianing Le, Long Chaudhari, Pratik Daniilidis, Kostas Wen, Chuan Jayaraman, Dinesh |
| author_facet | Shi, Junyao Yang, Rujia Chao, Kaitian Wan, Selina Bingqing Shao, Yifei Lei, Jiahui Qian, Jianing Le, Long Chaudhari, Pratik Daniilidis, Kostas Wen, Chuan Jayaraman, Dinesh |
| contents | Today's best-explored routes towards generalist robots center on collecting ever larger "observations-in actions-out" robotics datasets to train large end-to-end models, copying a recipe that has worked for vision-language models (VLMs). We pursue a road less traveled: building generalist policies directly around VLMs by augmenting their general capabilities with specific robot capabilities encapsulated in a carefully curated set of perception, planning, and control modules. In Maestro, a VLM coding agent dynamically composes these modules into a programmatic policy for the current task and scenario. Maestro's architecture benefits from a streamlined closed-loop interface without many manually imposed structural constraints, and a comprehensive and diverse tool repertoire. As a result, it largely surpasses today's VLA models for zero-shot performance on challenging manipulation skills. Further, Maestro is easily extensible to incorporate new modules, easily editable to suit new embodiments such as a quadruped-mounted arm, and even easily adapts from minimal real-world experiences through local code edits. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_00917 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Maestro: Orchestrating Robotics Modules with Vision-Language Models for Zero-Shot Generalist Robots Shi, Junyao Yang, Rujia Chao, Kaitian Wan, Selina Bingqing Shao, Yifei Lei, Jiahui Qian, Jianing Le, Long Chaudhari, Pratik Daniilidis, Kostas Wen, Chuan Jayaraman, Dinesh Robotics Artificial Intelligence Today's best-explored routes towards generalist robots center on collecting ever larger "observations-in actions-out" robotics datasets to train large end-to-end models, copying a recipe that has worked for vision-language models (VLMs). We pursue a road less traveled: building generalist policies directly around VLMs by augmenting their general capabilities with specific robot capabilities encapsulated in a carefully curated set of perception, planning, and control modules. In Maestro, a VLM coding agent dynamically composes these modules into a programmatic policy for the current task and scenario. Maestro's architecture benefits from a streamlined closed-loop interface without many manually imposed structural constraints, and a comprehensive and diverse tool repertoire. As a result, it largely surpasses today's VLA models for zero-shot performance on challenging manipulation skills. Further, Maestro is easily extensible to incorporate new modules, easily editable to suit new embodiments such as a quadruped-mounted arm, and even easily adapts from minimal real-world experiences through local code edits. |
| title | Maestro: Orchestrating Robotics Modules with Vision-Language Models for Zero-Shot Generalist Robots |
| topic | Robotics Artificial Intelligence |
| url | https://arxiv.org/abs/2511.00917 |