MAESTRO: Orchestrating Robotics Modules with Coding Agents for Zero-Shot Generalist Robots
MAESTRO: Orchestrating Robotics Modules with Coding Agents for Zero-Shot Generalist Robots
This research introduces MAESTRO, a system that leverages vision-language models as coding agents to write robot programs using a curated library of perception, planning, and control tools.
The paper proposes MAESTRO, a framework that uses vision-language models as coding agents to orchestrate robotics tool libraries for zero-shot generalist robot performance, contrasting this approach with dominant end-to-end policy training methods.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- code: Available Artifact link Open source S5
- weights: not_reported
- dataset: not_reported
- paper: not_reported
- code: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
arXiv:2511.00917v3 Announce Type: replace Abstract: Today's dominant route to generalist robots scales robotics data to train end-to-end policies. We instead use a vision-language model (VLM) as a coding agent that writes robot programs using a curated library of perception, planning, and control tools. MAESTRO's orchestration harness packages multi-step perception and action procedures into callable tools and guides
Open source S2
e to general-purpose robots alongside scaling end-to-end robot training data. Project site: https://maestro-robot.github.io. Code: https://github.com/penn-pal-lab/maestro.
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org