AIR-VLA+: Decoupling Movement and Manipulation via Cascaded Dual-Action Decoders with Asymmetric MoE for Aerial Robots

• URL: http://arxiv.org/abs/2606.12859v1
• Date: Thu, 11 Jun 2026 03:42:33 GMT
• Title: AIR-VLA+: Decoupling Movement and Manipulation via Cascaded Dual-Action Decoders with Asymmetric MoE for Aerial Robots
• Abstract summary: We propose a flow matching architecture specifically designed for aerial manipulation, featuring cascaded dual-action decoders and an asymmetric feature-level Mixture of Experts (MoE)<n>We construct cascaded manipulation and movement decoders, allowing the UAV to unidirectionally observe the manipulator's intent during movement to achieve workflow coordination.<n>Within the UAV movement decoder, we deploy an implicit MoE architecture, enabling different movement experts to spontaneously exhibit capacity inclinations for various task stages during training.
• Score: 21.904995550394933
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Aerial manipulation systems have long suffered from representation coupling in end-to-end control, as platform-level Unmanned Aerial Vehicle (UAV) movement and end-effector-level arm manipulation differ substantially in action scale, dynamics, and control objectives. In this paper, we propose AIR-VLA+, a flow matching action generation architecture specifically designed for aerial manipulation, featuring cascaded dual-action decoders and an asymmetric feature-level Mixture of Experts (MoE). We construct cascaded manipulation and movement decoders, allowing the UAV to unidirectionally observe the manipulator's intent during movement to achieve workflow coordination, while isolating the impact of UAV movement information backpropagation on arm manipulation stability. Addressing the characteristic that UAV movement is highly dependent on high-level semantics and responsible for task state transitions in aerial manipulation, we design an input feature enhancement module for the UAV movement decoder. This module introduces an implicit visual grasp projector to perceive the interaction state between the gripper and the object, and injects compressed global semantic features. Within the UAV movement decoder, we deploy an implicit MoE architecture, enabling different movement experts to spontaneously exhibit capacity inclinations for various task stages during training. Through dense soft blending computation on the feature manifold, the UAV movement is endowed with stronger task-stage adaptability. Experiments on the standardized AIR-VLA benchmark demonstrate that our method comprehensively surpasses all baselines with an overall average score of 48.0. The overall task completion score improves by 80.2\% compared to the single-head $π_{0.5}$ policy, effectively mitigating the heterogeneous coordinated control conflicts of composite robots.

Related papers

• Reinforcement Learning with Inner-loop Dynamics Estimator for Aerial Manipulation under Uncertainty [0.560928143828791]
  We present a hierarchical control framework that combines Reinforcement Learning (RL) with an inner-loop dynamics estimator to address this problem.<n>We validate the proposed approach on a custom quadrotor equipped with a 3-DoF manipulator through hardware experiments under varying payload conditions.
  arXiv  Detail & Related papers  (2026-06-15T12:16:01Z)
• VAIC: Vision-Guided Humanoid Agile Object Interaction Control via Decoupled Commands [56.029505206531155]
  Vision Guided Agile Interaction Control (VAIC) is a unified framework that bridges the gap between current controllers and humanoid robots.<n>First, a privileged teacher policy masters diverse interaction skills using precise object kinematics and exact environmental states.<n>Second, a deployable student policy distills these capabilities by replacing full body tracking with velocity targets across multiple axes.
  arXiv  Detail & Related papers  (2026-06-08T09:52:55Z)
• Aero-World: Action-Conditioned Aerial Video Generation from Inertial Controls [57.166020875486474]
  We propose textbfAero-World, a method for converting a pretrained image-to-video diffusion model into a controllable aerial video generator.<n>A frozen latent-space Physics Probe, trained independently on real video--IMU pairs, provides differentiable inertial-consistency supervision during LoRA finetuning.<n>On AeroBench, Aero-World improves mean AAS from 57.7 to 63.6 over action-only finetuning and gives a stronger quality-control trade-off than AirScape.
  arXiv  Detail & Related papers  (2026-05-19T12:02:17Z)
• ULTRA: Unified Multimodal Control for Autonomous Humanoid Whole-Body Loco-Manipulation [55.467742403416175]
  We introduce a physics-driven neural algorithm that translates large-scale motion capture to humanoid embodiments.<n>We learn a unified multimodal controller that supports both dense references and sparse task specifications.<n>Results show that ULTRA generalizes to autonomous, goal-conditioned whole-body loco-manipulation from egocentric perception.
  arXiv  Detail & Related papers  (2026-03-03T18:59:29Z)
• Glove2UAV: A Wearable IMU-Based Glove for Intuitive Control of UAV [1.9044180551673717]
  Glove2UAV is a wearable IMU-glove interface for intuitive UAV control through hand and finger gestures.<n>It streams inertial measurements in real time and estimates palm and finger orientations.<n>Vibrotactile feedback is triggered when flight speed exceeds predefined threshold values.
  arXiv  Detail & Related papers  (2026-01-22T09:03:57Z)
• PreGME: Prescribed Performance Control of Aerial Manipulators based on Variable-Gain ESO [12.679195185695164]
  We propose a novel prescribed performance motion control framework based on variable-gain extended state observers (ESOs)<n>The method includes variable-gain ESOs for real-time estimation of dynamic coupling and a prescribed performance flight control.<n> Experimental results demonstrate that even under the dynamic coupling caused by rapid robotic arm motion, the proposed method achieves high tracking performance.
  arXiv  Detail & Related papers  (2025-12-28T14:54:06Z)
• Robotic VLA Benefits from Joint Learning with Motion Image Diffusion [114.60268819583017]
  Vision-Language-Action (VLA) models have achieved remarkable progress in robotic manipulation by mapping multimodal observations and instructions directly to actions.<n>We propose joint learning with motion image diffusion, a novel strategy that enhances VLA models with motion reasoning capabilities.<n> Experiments in both simulation and real-world environments demonstrate that joint learning with motion image diffusion improves the success rate of pi-series VLAs to 97.5%.
  arXiv  Detail & Related papers  (2025-12-19T19:07:53Z)
• Tracking the Unstable: Appearance-Guided Motion Modeling for Robust Multi-Object Tracking in UAV-Captured Videos [58.156141601478794]
  Multi-object tracking (UAVT) aims to track multiple objects while maintaining consistent identities across frames of a given video.<n>Existing methods typically model motion cues and appearance separately, overlooking their interplay and resulting in suboptimal tracking performance.<n>We propose AMOT, which exploits appearance and motion cues through two key components: an Appearance-Motion Consistency (AMC) matrix and a Motion-aware Track Continuation (MTC) module.
  arXiv  Detail & Related papers  (2025-08-03T12:06:47Z)
• LLM Meets the Sky: Heuristic Multi-Agent Reinforcement Learning for Secure Heterogeneous UAV Networks [57.27815890269697]
  This work focuses on maximizing the secrecy rate in heterogeneous UAV networks (HetUAVNs) under energy constraints.<n>We introduce a Large Language Model (LLM)-guided multi-agent learning approach.<n>Results show that our method outperforms existing baselines in secrecy and energy efficiency.
  arXiv  Detail & Related papers  (2025-07-23T04:22:57Z)
• Actuator Trajectory Planning for UAVs with Overhead Manipulator using Reinforcement Learning [0.3222802562733786]
  We develop a UAV equipped with a controllable arm with two degrees of freedom to carry out actuation tasks on the fly. Our solution is based on employing a Q-learning method to control the trajectory of the tip of the arm, also called end-effector. Our method achieves 92% accuracy in terms of average displacement error using Q-learning with 15,000 episodes.
  arXiv  Detail & Related papers  (2023-08-24T15:06:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.