Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.21168 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917546423222272 |
|---|---|
| author | Ruan, Qiyu Wang, Yuxuan Li, He Li, Zhenning Xu, Cheng-zhong |
| author_facet | Ruan, Qiyu Wang, Yuxuan Li, He Li, Zhenning Xu, Cheng-zhong |
| contents | Safety-critical scenarios are central to evaluating autonomous driving systems, yet their rarity in naturalistic logs makes simulation-based stress testing indispensable. Most scenario generation methods treat surrounding agents as adversaries, but they either (i) induce failures without explicitly modeling vehicle-road physical limits, yielding visually extreme yet physically unsolvable crashes, or (ii) enforce physical feasibility or policy feasibility in isolation, which can over-focus on aggressive maneuvers or remain tied to a controller-dependent capability boundary. We propose ScenePilot, a feasibility-guided, boundary-driven framework that targets the boundary band: scenarios that are physically solvable in principle yet still cause the deployed autonomy stack to fail. We formulate generation as constrained multi-objective reinforcement learning, combining an RSS-derived physical-feasibility score $σ$ with an online-learned AV-risk predictor $Φ$, and introduce step-level feasibility-aware shielding to keep exploration near the feasibility boundary while avoiding infeasible artifacts. Experiments on SafeBench with multiple planners show that ScenePilot yields substantially higher collision rates (+6.2 percentage points) while preserving physical validity, and that adversarial fine-tuning on these boundary-band scenarios consistently reduces downstream crash rates. The code is available at https://github.com/QiyuRuan/ScenePilot. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_21168 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | ScenePilot: Controllable Boundary-Driven Critical Scenario Generation for Autonomous Driving Ruan, Qiyu Wang, Yuxuan Li, He Li, Zhenning Xu, Cheng-zhong Artificial Intelligence Safety-critical scenarios are central to evaluating autonomous driving systems, yet their rarity in naturalistic logs makes simulation-based stress testing indispensable. Most scenario generation methods treat surrounding agents as adversaries, but they either (i) induce failures without explicitly modeling vehicle-road physical limits, yielding visually extreme yet physically unsolvable crashes, or (ii) enforce physical feasibility or policy feasibility in isolation, which can over-focus on aggressive maneuvers or remain tied to a controller-dependent capability boundary. We propose ScenePilot, a feasibility-guided, boundary-driven framework that targets the boundary band: scenarios that are physically solvable in principle yet still cause the deployed autonomy stack to fail. We formulate generation as constrained multi-objective reinforcement learning, combining an RSS-derived physical-feasibility score $σ$ with an online-learned AV-risk predictor $Φ$, and introduce step-level feasibility-aware shielding to keep exploration near the feasibility boundary while avoiding infeasible artifacts. Experiments on SafeBench with multiple planners show that ScenePilot yields substantially higher collision rates (+6.2 percentage points) while preserving physical validity, and that adversarial fine-tuning on these boundary-band scenarios consistently reduces downstream crash rates. The code is available at https://github.com/QiyuRuan/ScenePilot. |
| title | ScenePilot: Controllable Boundary-Driven Critical Scenario Generation for Autonomous Driving |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2605.21168 |