CRASH: Challenging Reinforcement-Learning Based Adversarial Scenarios For Safety Hardening

TL;DR

CRASH is an adversarial reinforcement learning framework that generates challenging traffic scenarios to test and improve autonomous vehicle safety, significantly increasing failure detection and reducing collision rates.

Contribution

The paper introduces CRASH, a novel RL-based adversarial scenario generator, and a safety hardening method that iteratively enhances AV motion planners using adversarial failures.

Findings

CRASH can induce over 90% collision rates in tested scenarios.

Safety hardening reduces AV collision rates by 26%.

The approach effectively falsifies both rule-based and learning-based planners.

Abstract

Ensuring the safety of autonomous vehicles (AVs) requires identifying rare but critical failure cases that on-road testing alone cannot discover. High-fidelity simulations provide a scalable alternative, but automatically generating realistic and diverse traffic scenarios that can effectively stress test AV motion planners remains a key challenge. This paper introduces CRASH - Challenging Reinforcement-learning based Adversarial scenarios for Safety Hardening - an adversarial deep reinforcement learning framework to address this issue. First CRASH can control adversarial Non Player Character (NPC) agents in an AV simulator to automatically induce collisions with the Ego vehicle, falsifying its motion planner. We also propose a novel approach, that we term safety hardening, which iteratively refines the motion planner by simulating improvement scenarios against adversarial agents, leveraging the failure cases to strengthen the AV stack. CRASH is evaluated on a simplified two-lane highway scenario, demonstrating its ability to falsify both rule-based and learning-based planners with collision rates exceeding 90%. Additionally, safety hardening reduces the Ego vehicle's collision rate by 26%. While preliminary, these results highlight RL-based safety hardening as a promising approach for scenario-driven simulation testing for autonomous vehicles.