ProDrive: Proactive Planning for Autonomous Driving via Ego-Environment Co-Evolution

arXiv 2026 54.3 method, application

TLDR

ProDrive uses a BEV world model and ego-environment co-evolution for proactive autonomous driving planning, outperforming baselines on NAVSIM v1.

Reasoning

The paper presents a novel end-to-end framework coupling a trajectory planner with a world model for proactive planning, which is a strength. However, it only evaluates on a simulation benchmark (NAVSIM v1) without real-world experiments, and the abstract lacks discussion of limitations or failure cases.

Read-first score

Read-first score 54.3, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 43.

Recency 6%

100

Uses a gentle age decay so recent papers surface without erasing older foundations. 2026

Methodology quality 18%

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

Citation impact 18%

61.9

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. citation_normalized_percentile=0.61947566

Topical relevance 29%

61.4

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

Reproducibility 18%

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

Citation velocity 12%

Citation velocity estimates citations per publication-year to reduce old-paper bias. velocity=0.00

Field roles

FrontierBridgeMethodology anchor

Rank sensitivity

Stability: volatile; rank range: 266.

Keyword Scores

world model

interactive world model

world dynamics prediction

generative world model

model-based reinforcement learning world model

world simulator

video world model

Deep Analysis

Innovations

Proactive planning via ego-environment co-evolution with bidirectional coupling
Joint end-to-end training of a query-centric trajectory planner and a BEV world model
Planning-aware ego tokens and parallel candidate evaluation preserving gradient flow

Methodology

ProDrive jointly trains a query-centric trajectory planner and a bird's-eye-view (BEV) world model end-to-end. The planner generates diverse candidate trajectories and planning-aware ego tokens, while the world model predicts future scene evolution conditioned on them. By injecting planner features into the world model and evaluating all candidates in parallel, the framework preserves end-to-end gradient flow and allows future outcome assessment to directly shape planning.

Key Results

On NAVSIM v1, ProDrive outperforms strong baselines in both safety and planning efficiency, and ablations validate the effectiveness of the proposed ego-environment coupling design.

Resources

Paper