FlowDrive: Energy Flow Field for End-to-End Autonomous Driving

2025-09-17

AI Summary

Existing end-to-end autonomous driving methods rely on implicit bird’s-eye-view (BEV) features and lack explicit modeling of risk and guidance priors, resulting in insufficient planning safety and interpretability. To address this, we propose the Energy Flow Field (EFF) framework: it explicitly constructs a risk potential field and a lane-attracting field in BEV space, endowing trajectory planning with physical interpretability; introduces a feature-level gating mechanism to decouple motion intent prediction from trajectory denoising, enhancing multimodal planning diversity and robustness; and integrates BEV semantic prior encoding with a conditional diffusion-based planner for high-fidelity trajectory generation. Evaluated on the NAVSIM v2 benchmark, EFF achieves 86.3 EPDMS—significantly outperforming prior methods—while simultaneously ensuring safety, interpretability, and planning quality.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Intelligent Robots: Motion and Path Planning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments

Abstract

Recent advances in end-to-end autonomous driving leverage multi-view images to construct BEV representations for motion planning. In motion planning, autonomous vehicles need considering both hard constraints imposed by geometrically occupied obstacles (e.g., vehicles, pedestrians) and soft, rule-based semantics with no explicit geometry (e.g., lane boundaries, traffic priors). However, existing end-to-end frameworks typically rely on BEV features learned in an implicit manner, lacking explicit modeling of risk and guidance priors for safe and interpretable planning. To address this, we propose FlowDrive, a novel framework that introduces physically interpretable energy-based flow fields-including risk potential and lane attraction fields-to encode semantic priors and safety cues into the BEV space. These flow-aware features enable adaptive refinement of anchor trajectories and serve as interpretable guidance for trajectory generation. Moreover, FlowDrive decouples motion intent prediction from trajectory denoising via a conditional diffusion planner with feature-level gating, alleviating task interference and enhancing multimodal diversity. Experiments on the NAVSIM v2 benchmark demonstrate that FlowDrive achieves state-of-the-art performance with an EPDMS of 86.3, surpassing prior baselines in both safety and planning quality. The project is available at https://astrixdrive.github.io/FlowDrive.github.io/.

Problem

• Modeling risk and guidance priors for safe autonomous driving
• Integrating semantic and safety cues into BEV representations
• Decoupling motion intent prediction from trajectory denoising

Innovation

• Energy-based flow fields for semantic priors
• Conditional diffusion planner with gating
• Adaptive refinement of anchor trajectories