Skip to content

Aero-VPA: Few-Step Flow-Matching Control over Non-Reconstructive JEPA Latents for Agile Flight on Edge Hardware

Nov 2026 · 0 citations · 76 references
Computer Science

TL;DR

This version reports the first closed-loop evaluation: PX4 software-in-the-loop simulation of quadrotor missions in procedurally generated forests, and presents Aero-VPA, a lightweight Vision-Predictor-Action architecture for 3D navigation on edge hardware.

Abstract

Vision-Language-Action (VLA) models generalize across tasks, but their autoregressive decoders need sequential forward passes in proportion to the action horizon, each bounded below on embedded aerial platforms by streaming billions of parameters from memory. We present Aero-VPA, a lightweight Vision-Predictor-Action architecture for 3D navigation on edge hardware. A compact RGB-D Vision Transformer and a five-head ensemble predictor, trained without pixel reconstruction, forecast the next latent state; a Conditional Flow Matching solver generates a velocity and yaw-rate chunk in two Euler steps, so each decision takes five sequential network evaluations; and the ensemble dispersion drives an uncertainty fallback. Version 1 specified the architecture, deployment and protocol without results. This version reports the first closed-loop evaluation: PX4 software-in-the-loop simulation of quadrotor missions in procedurally generated forests. Behaviour cloning failed on all 60 missions despite open-loop errors of 0.02-0.04 m/s: the policy copies its last command, a copycat form of causal confusion. DAgger with a privileged, safety-gated supervisor gave the first successes but plateaued at 11.1-21.1% across six checkpoints. A map-free runtime shield with a 15 s depth memory, barrier constraints, a cap on motion outside the 73-degree field of view and backoff along the flown path raised the first DAgger policy's success on 180 paired missions from 21.1% to 52.2% (95% CI [45.0, 59.4]; exact McNemar p = 1.2 x 10^-10) and cut its crashes from 72.2% to 22.2%, while timeouts rose from 2.8% to 23.3%. All six checkpoints gained 30.0-34.4 percentage points, at mean speeds of 1.16-1.36 m/s. The code (https://github.com/ylee-sling/Aero-VPA), the trained policy (72.6 MB, https://huggingface.co/ylee-sling/aero-vpa) and the full benchmark dataset are public.

View source

Similar papers

Open access Oct 2026

MSTR-PPO: A Multi-Scale Temporal Residual PPO Controller for Freeze-Prone Indoor Navigation

Local navigation controllers often stop making progress in constrained indoor environments, where narrow geometry and moving obstacles compress the feasible motion space until the robot stalls and falls back on repeated recovery behaviors. We treat such freezing as a failure of local control arising from three observ...

Jin-Long Chen, Yang-Ying Kou, Wu-Jun Che et al. · 0 citations
#artificial intelligence Preprint Oct 2026

X-Fly: Learning to Reach and Stop in Full-Chunk Aerial Vision-Language Navigation

Aerial vision-language navigation (VLN) agents built on large vision-language models are expensive to query. Executing a complete predicted action chunk before replanning amortizes this cost, reducing model inference from 195 to 42 ms per executed waypoint for our policy. It also moves the states at which the policy is...

Rui-Bin Ju, Rui Cai, Min Xu et al. · 1 citation
Sep 2026

VLM-MPPI: Grounding Natural Language in Behaviorally Diverse Trajectories for Aerial Navigation

A hierarchical UAV navigation framework that aligns natural-language intent with dynamically feasible flight behaviors in cluttered indoor environments is presented, and a parallelized ensemble of six behavior-conditioned Model Predictive Path Integral (MPPI) planners are employed.

Hanbing Zhang, Fang-Guo Zhao, Ze-Rui Li et al. · 0 citations
Preprint Oct 2026

Leap-VLA: High-Rate Robot Control from a Low-Rate VLM Backbone

Vision-language-action (VLA) models re-encode the full observation through a multi-billion-parameter vision-language model (VLM) at every control step, even though scene semantics evolve at roughly 1 Hz while reactive control requires 20-50 Hz. We present a VLA that decouples the two: a VLM runs as a low-frequency sema...

Ling-Chu Yang, Yun-Jie Shen, Zhe-Yu Li et al. · 0 citations
Sep 2026

C2Nav: Compare Before You Commit for Zero-Shot Vision-and-Language Navigation

This work instantiates a complementary model-robot interface: the VLM compares controller-constructed alternatives, while geometry, thresholds, action magnitude, and execution remain on the physical side in C2Nav, a training-free framework with three coordinated faculties.

Run-Tian Zheng, Cong-Peng Zhang, Ying Liu · 1 citation
Preprint Oct 2026

End-to-End Vision-Language-Action Policies for Autonomous Lunar Rover Docking

A vision-language-action policy for autonomous lunar rover docking and its performance relative to a modular navigation pipeline is designed and a scalable training framework is developed that generates safety-aware expert demonstrations in a high-fidelity lunar simulation environment.

Si-Tan Huang, G. Gao · 0 citations

Related blog posts

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.