Aligning Cyber Space with Physical World: A Comprehensive Survey on Embodied AI

TMECH 25 2025 56.3 survey

TLDR

A comprehensive survey on Embodied AI, covering perception, interaction, agents, sim-to-real, and the role of MLMs and World Models.

Reasoning

Strengths: Broad coverage of recent advances in Embodied AI, including Multi-modal Large Models and World Models. Weaknesses: Lacks novel experiments or empirical evaluations; as a survey, it does not provide new results or real-world benchmarks.

Read-first score

Read-first score 56.3, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 19.

Recency 8%

86.7

Uses a gentle age decay so recent papers surface without erasing older foundations. 2025

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=True; dataset=False; markers=dataset,github

Methodology quality 25%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=analysis,dataset

Topical relevance 42%

27.1

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Field roles

FrontierMethodology anchorReproducibility anchor

Rank sensitivity

Stability: volatile; rank range: 646.

Keyword Scores

world model

world simulator

generative world model

interactive world model

model-based reinforcement learning world model

video world model

world dynamics prediction

Deep Analysis

Innovations

Comprehensive survey covering four main research targets: embodied perception, embodied interaction, embodied agent, and sim-to-real adaptation
Analysis of Multi-modal Large Models (MLMs) and World Models (WMs) as promising architectures for embodied agents
Identification of challenges and future directions for Embodied AI

Methodology

This survey provides a comprehensive exploration of Embodied AI by first reviewing representative embodied robots and simulators, then analyzing four main research targets (embodied perception, interaction, agent, sim-to-real adaptation) covering state-of-the-art methods, essential paradigms, and comprehensive datasets. It also explores the complexities of Multi-modal Large Models in virtual and real embodied agents.

Key Results

The survey presents a structured overview of the latest advancements in Embodied AI, highlighting the significance of MLMs and WMs for bridging cyberspace and the physical world, and summarizes challenges and limitations of the field.

Limitations

The abstract does not enumerate specific limitations; they are summarized within the paper.

Resources

Paper Code