# Dance of the ADS: Orchestrating Failures through Historically-Informed Scenario Fuzzing

arXiv: https://arxiv.org/abs/2407.04359 · HTML: https://mlsanity.com/paper/2407.04359/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 45 references that are in the map.
- First posted on arXiv 5 Jul 2024.

## Builds on (most cited first, 18 of 18)

- [CARLA: An Open Urban Driving Simulator](https://mlsanity.com/paper/1711.03938/) (2017)
- [NEAT: Neural Attention Fields for End-to-End Autonomous Driving](https://mlsanity.com/paper/2109.04456/) (2021)
- [LGSVL Simulator: A High Fidelity Simulator for Autonomous Driving](https://mlsanity.com/paper/2005.03778/) (2020)
- [A Survey on Safety-Critical Driving Scenario Generation -- A Methodological Perspective](https://mlsanity.com/paper/2202.02215/) (2022)
- [KING: Generating Safety-Critical Driving Scenarios for Robust Imitation via Kinematics Gradients](https://mlsanity.com/paper/2204.13683/) (2022)
- [On Adversarial Robustness of Trajectory Prediction for Autonomous Vehicles](https://mlsanity.com/paper/2201.05057/) (2022)
- [SafeBench: A Benchmarking Platform for Safety Evaluation of Autonomous Vehicles](https://mlsanity.com/paper/2206.09682/) (2022)
- [Ontology based Scene Creation for the Development of Automated Vehicles](https://mlsanity.com/paper/1704.01006/) (2017)
- [A Survey on Automated Driving System Testing: Landscapes and Trends](https://mlsanity.com/paper/2206.05961/) (2022)
- [A Survey on Scenario-Based Testing for Automated Driving Systems in High-Fidelity Simulation](https://mlsanity.com/paper/2112.00964/) (2021)
- [CausalAF: Causal Autoregressive Flow for Safety-Critical Driving Scenario Generation](https://mlsanity.com/paper/2110.13939/) (2021)
- [Neural Network Guided Evolutionary Fuzzing for Finding Traffic Violations of Autonomous Vehicles](https://mlsanity.com/paper/2109.06126/) (2021)
- [DriveFuzz: Discovering Autonomous Driving Bugs through Driving Quality-Guided Fuzzing](https://mlsanity.com/paper/2211.01829/) (2022)
- [BehAVExplor: Behavior Diversity Guided Testing for Autonomous Driving Systems](https://mlsanity.com/paper/2307.07493/) (2023)
- [ANTI-CARLA: An Adversarial Testing Framework for Autonomous Vehicles in CARLA](https://mlsanity.com/paper/2208.06309/) (2022)
- [Intersection focused Situation Coverage-based Verification and Validation Framework for Autonomous Vehicles Implemented in CARLA](https://mlsanity.com/paper/2112.14706/) (2021)
- [Risk-Aware Scene Sampling for Dynamic Assurance of Autonomous Systems](https://mlsanity.com/paper/2202.13510/) (2022)
- [Critical concrete scenario generation using scenario-based falsification](https://mlsanity.com/paper/2208.13329/) (2022)

## Built on by (most cited first, 2 of 2)

- [ROCAS: Root Cause Analysis of Autonomous Driving Accidents via Cyber-Physical Co-mutation](https://mlsanity.com/paper/2409.07774/) (2024)
- [ROMAN: Reward-Orchestrated Multi-Head Attention Network for Autonomous Driving System Testing](https://mlsanity.com/paper/2602.05629/) (2026)

## Abstract (arXiv)

As autonomous driving systems (ADS) advance towards higher levels of autonomy, orchestrating their safety verification becomes increasingly intricate. This paper unveils ScenarioFuzz, a pioneering scenario-based fuzz testing methodology. Designed like a choreographer who understands the past performances, it uncovers vulnerabilities in ADS without the crutch of predefined scenarios. Leveraging map road networks, such as OPENDRIVE, we extract essential data to form a foundational scenario seed corpus. This corpus, enriched with pertinent information, provides the necessary boundaries for fuzz testing in the absence of starting scenarios. Our approach integrates specialized mutators and mutation techniques, combined with a graph neural network model, to predict and filter out high-risk scenario seeds, optimizing the fuzzing process using historical test data. Compared to other methods, our approach reduces the time cost by an average of 60.3%, while the number of error scenarios discovered per unit of time increases by 103%. Furthermore, we propose a self-supervised collision trajectory clustering method, which aids in identifying and summarizing 54 high-risk scenario categories prone to inducing ADS faults. Our experiments have successfully uncovered 58 bugs across six tested systems, emphasizing the critical safety concerns of ADS.