World-Action Models for Robot Learning and Control: A Survey

2026-09-13

AI Summary

本文综述了世界-动作模型（WAMs），旨在通过结合未来状态预测与可执行动作生成，解决机器人在开放环境中面对的部分可观测性、物理约束及动态任务背景下的学习和控制问题。

Abstract

Robots operating in open environments act under partial observability, physical constraints, and dynamic task contexts. Beyond mapping observations and language instructions to actions, they must anticipate how candidate actions may affect future states and task-relevant outcomes. Recent advances in world models, video generation, and Vision-Language-Action (VLA) policies have motivated the development of World-Action Models (WAMs), which couple future world prediction with executable action generation. This survey provides a robotics-oriented review of WAMs. We clarify their scope relative to conventional world models, model-based reinforcement learning, action-conditioned video generation, and reactive VLA policies, and organize existing methods through a unified taxonomy covering representations, transition modeling, action interfaces, architectures, training pipelines, data modalities, and scaling strategies. We further review applications of WAMs in manipulation, navigation, and autonomous driving, and we summarize the datasets, benchmarks, metrics, and protocols used to evaluate WAM systems. Finally, we discuss key challenges in action alignment, world-action factorization, spatial and multi-view consistency, long-horizon memory, neural simulation for closed-loop policy learning, and efficient inference. Taken together, this survey aims to provide a concise technical foundation for integrating predictive world modeling with action generation, toward more reliable embodied robot intelligence. Project page: https://rcl-robotics.github.io/Awesome-World-Action-Models.

Problem

• partial observability
• physical constraints
• dynamic task contexts
• future state prediction
• task-relevant outcomes

Innovation

• World-Action Models
• future world prediction
• action generation
• robotics applications
• predictive modeling