World Model for Robot Learning: A Comprehensive Survey

2026-04-30

AI Summary

This work presents the first comprehensive survey of world models in robot learning, offering a systematic synthesis of their key paradigms, functional roles, and evolutionary trajectory. Addressing the current fragmentation across architectures and application domains, the study elucidates how world models interface with policy learning, serve as learnable simulators, and evolve from imagination-based generation toward structured control and foundation model integration. Through an integrative analysis spanning predictive modeling, reinforcement learning, video generation, navigation, and autonomous driving, the paper clarifies the core contributions of world models, catalogs representative datasets, benchmarks, and evaluation protocols, and establishes a continuously updated open-source repository to provide both a foundational reference and infrastructural support for future research in this rapidly advancing field.

Abstract

World models, which are predictive representations of how environments evolve under actions, have become a central component of robot learning. They support policy learning, planning, simulation, evaluation, data generation, and have advanced rapidly with the rise of foundation models and large-scale video generation. However, the literature remains fragmented across architectures, functional roles, and embodied application domains. To address this gap, we present a comprehensive review of world models from a robot-learning perspective. We examine how world models are coupled with robot policies, how they serve as learned simulators for reinforcement learning and evaluation, and how robotic video world models have progressed from imagination-based generation to controllable, structured, and foundation-scale formulations. We further connect these ideas to navigation and autonomous driving, and summarize representative datasets, benchmarks, and evaluation protocols. Overall, this survey systematically reviews the rapidly growing literature on world models for robot learning, clarifies key paradigms and applications, and highlights major challenges and future directions for predictive modeling in embodied agents. To facilitate continued access to newly emerging works, benchmarks, and resources, we will maintain and regularly update the accompanying GitHub repository alongside this survey.

Problem

• world models
• robot learning
• predictive modeling
• embodied agents
• literature fragmentation

Innovation

• World Models
• Robot Learning
• Learned Simulators
• Video Generation
• Embodied Agents