AIROMay 2, 2025

Seeking to Collide: Online Safety-Critical Scenario Generation for Autonomous Driving with Retrieval Augmented Large Language Models

arXiv:2505.00972v27 citationsh-index: 16
Originality Incremental advance
AI Analysis

This addresses the need for exposing rare, safety-critical corner cases in autonomous driving simulation, though it is incremental as it builds on existing LLM and retrieval methods.

The paper tackles the problem of generating safety-critical driving scenarios for autonomous vehicle testing by introducing an online, retrieval-augmented LLM framework, which reduces the mean minimum time-to-collision from 1.62 to 1.08 seconds and achieves a 75% collision rate in evaluations.

Simulation-based testing is crucial for validating autonomous vehicles (AVs), yet existing scenario generation methods either overfit to common driving patterns or operate in an offline, non-interactive manner that fails to expose rare, safety-critical corner cases. In this paper, we introduce an online, retrieval-augmented large language model (LLM) framework for generating safety-critical driving scenarios. Our method first employs an LLM-based behavior analyzer to infer the most dangerous intent of the background vehicle from the observed state, then queries additional LLM agents to synthesize feasible adversarial trajectories. To mitigate catastrophic forgetting and accelerate adaptation, we augment the framework with a dynamic memorization and retrieval bank of intent-planner pairs, automatically expanding its behavioral library when novel intents arise. Evaluations using the Waymo Open Motion Dataset demonstrate that our model reduces the mean minimum time-to-collision from 1.62 to 1.08 s and incurs a 75% collision rate, substantially outperforming baselines.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes