Vehicle Dynamics Embedded World Models for Autonomous Driving

arXiv 25.12 2025 48.1 method, application

TLDR

Proposes VDD, a world model that decouples ego-vehicle dynamics from environment dynamics for robust autonomous driving in simulation.

Reasoning

The paper clearly addresses a key limitation in existing world models for autonomous driving by separating vehicle and environment dynamics, and introduces practical training/deployment strategies. However, it lacks real-world validation and does not discuss generative or interactive aspects, limiting its scope.

Read-first score

Read-first score 48.1, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 35.

Recency 8%

86.7

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

Topical relevance 42%

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

Methodology quality 25%

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

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=none

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 434.

Keyword Scores

world model

world dynamics prediction

model-based reinforcement learning world model

world simulator

generative world model

interactive world model

video world model

Deep Analysis

Innovations

Decoupling ego-vehicle dynamics from environmental transition dynamics in world models
Policy Adjustment during Deployment (PAD) strategy
Policy Augmentation during Training (PAT) strategy

Methodology

The paper proposes the Vehicle Dynamics embedded Dreamer (VDD) method, which decouples the modeling of ego-vehicle dynamics from environmental transition dynamics. It introduces two strategies: Policy Adjustment during Deployment (PAD) and Policy Augmentation during Training (PAT) to enhance policy robustness. The model is evaluated in simulated environments against existing approaches.

Key Results

The proposed VDD method significantly improves driving performance and robustness to variations in vehicle dynamics, outperforming existing approaches in comprehensive simulated experiments.

Resources

Paper