Planning-Oriented End-to-End Autonomous Driving: Architectures, Evaluation, and Emerging Paradigms

2026-08-20

AI Summary

该论文综述了端到端自动驾驶技术从直接控制回归向规划导向系统的转变，探讨了使用结构化表示、轨迹级输出及更真实评估协议的方法。

Abstract

End-to-end autonomous driving has evolved from camera-to-control regression toward planning-oriented systems that use structured representations, trajectory-level outputs, and increasingly realistic evaluation protocols. This survey reviews this transition across behavior cloning, conditional imitation learning, privileged distillation, BEV and vectorized planning, unified perception-prediction-planning architectures, world-model-based planners, and vision-language-action systems. We argue that the key distinction in modern end-to-end driving is not whether intermediate representations are used, but whether they are learned, supervised, and evaluated to support safe, feasible, and route-compliant planning. To organize the literature, we synthesize existing methods along four axes: input representation, planning output, supervision signal, and evaluation protocol. We further examine the benchmark shift from open-loop trajectory matching to closed-loop simulation, non-reactive real-log evaluation, long-tail testing, and human-preference-aware metrics. Our analysis highlights that architectural progress is difficult to interpret without benchmark-consistent evaluation, and that displacement-based open-loop metrics alone provide limited evidence for safe and human-aligned driving. We conclude with open challenges in uncertainty-aware planning, learner-expert mismatch, runtime safety assurance, language-action grounding, world-model validation, and reproducible benchmarking.

Problem

• end-to-end autonomous driving
• planning-oriented systems
• structured representations
• trajectory-level outputs
• evaluation protocols

Innovation

• planning-oriented systems
• structured representations
• trajectory-level outputs
• realistic evaluation protocols
• benchmark-consistent evaluation