Create

Sign in to ReadmeX

or

How we handle your data: Privacy Policy

ReadmeX

Your AI reading companion

What would you like to catch up on?

Pick a briefing, or ask your own question.

Your community and people briefings stay personal to you.

Story

OpenFlyV2 runs vision-language-action control onboard a drone

AI summary

A paper submitted to arXiv on 29 October 2026 presents OpenFlyV2, a unified onboard vision-language-action architecture for aerial perception and navigation. It reports 10 Hz onboard control on an NVIDIA Jetson Orin NX, generating 12-waypoint action chunks, and a 48% navigation success rate across four real-world environments without real-world aerial training data. The authors also describe a simulation dataset with approximately 70,000 navigation trajectories and 60,000 vision-language instruction-response pairs.

In-depth insight

The paper reports a single model combining language-grounded perception, spatial reasoning, and continuous flight control on resource-constrained hardware, while transferring from simulation to physical quadrotors without real-world aerial training data.

Computer Science > Robotics

arXiv:2610.39287 (cs)

[Submitted on 29 Oct 2026]

Title:OpenFlyV2: An onboard vision-language-action model for aerial perception and navigation

Authors:Junli Liu, Zhongrui You, Chenhui Li, Zhen Li, Yuyang Tian, Qizhen Chen, Delin Qu, Haoming Song, Dong Wang, Xuelong Li, Zhigang Wang, Bin Zhao

View a PDF of the paper titled OpenFlyV2: An onboard vision-language-action model for aerial perception and navigation, by Junli Liu and 11 other authors

#OpenFlyV2#NVIDIA Jetson Orin NX#NVIDIA

Original · 3 min readRead the original
Group chat

No comments yet. Start the conversation.