Vision Language Action (VLA) Models for Unmanned Aerial Robotics and Bimanual Manipulation: A Review

2026-07-07

AI Summary

This study addresses the development of a unified framework for vision–language–action (VLA) models applied to both bimanual robotic manipulation and autonomous drone control. Synthesizing insights from 183 studies published between 2017 and 2026, the work provides a comprehensive analysis across seven dimensions—including architectural design, training paradigms, action representations, and language alignment. It reveals, for the first time, significant commonalities between the two platforms in coordination strategies, training methodologies, and action encoding schemes, and articulates fourteen cross-domain research directions. The findings demonstrate that VLA has emerged as the dominant paradigm for learning-based robotic manipulation, exhibiting strong transfer efficacy in both high-degree-of-freedom dual-arm systems and resource-constrained unmanned aerial vehicles.

Abstract

Vision Language Action (VLA) models unify visual perception, natural-language understanding, and action generation within a single foundation model, allowing a robot to follow instructions such as fold the towel or fly to the red building directly from camera images. Because VLAs inherit world knowledge from internet-scale pre-training, they have become the dominant framework for learning-based manipulation, with bimanual coordination serving as the most demanding testbed: two arms with 7 degrees of freedom each must move in concert to fold, assemble, and reorient objects. Unmanned aerial robotics faces a structurally similar challenge: a drone must coordinate thrust, attitude, and increasingly gripper commands from visual observations under strict latency and payload constraints. This review covers 183 contributions spanning 2017-2026 and organized along seven dimensions: VLA architectures, training recipes, action representations, bimanual coordination (2022-2026), unmanned aerial vehicle (UAV) navigation and control (2017-2026), language grounding, and cross-cutting concerns including memory and world models. We show that the coordination strategies, training recipes, and action representations developed for bimanual VLAs transfer to unmanned aerial systems and identify fourteen research directions across both domains.

Problem

• Vision Language Action
• bimanual manipulation
• unmanned aerial robotics
• action coordination
• multimodal control

Innovation

• Vision Language Action
• bimanual manipulation
• unmanned aerial robotics
• action representation
• cross-domain transfer