الملخص
Existing adaptive reasoning methods for driving Vision-Language-Action (VLA) models primarily focus on whether to reason, overlooking how reasoning should differ across driving situations. Our key insight is that while scene complexity informs reasoning depth, dynamic risk is equally critical for deciding how to reason in time-critical situations. We therefore propose CAR-VLA, a unified driving VLA model that jointly considers scene complexity and dynamic risk to guide reasoning depth, urgency, and focus. CAR-VLA maps four complexity--risk categories to three reasoning modes: \textit{Fast Intuition} for direct trajectory generation in simple low-risk scenes, \textit{Slow Thinking} for deliberate reasoning in complex low-risk scenes, and \textit{Reflex Response} for compact, hazard-focused reasoning in high-risk scenes regardless of complexity. Rather than merely shortening deliberation, Reflex Response centers reasoning on the most critical hazard and the immediate safe response. We train CAR-VLA through progressive supervised learning that links scene assessment, reasoning-mode selection, and trajectory generation, followed by reasoning-augmented reinforcement learning to improve driving quality and reasoning behavior. Experiments on NAVSIM v1(91.1 PDMS), NAVSIM v2(90.3 EPDMS), and Navhard(35.0 EPDMS) demonstrate competitive driving performance. Qualitative comparisons on navtest and in-house high-risk scenarios further illustrate risk-aware reasoning and hazard-responsive trajectory generation. The code for this paper will be released publicly at: https://github.com/chenxl124578/CAR-VLA.git
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Chen, X., He, Z., Liang, Y., Li, X., Chen, H., Shi, F., Zhao, M., Meng, W., Chen, Z., Zhu, Z., Jin, Z., Wang, H., Wang, Q., Liang, J., Li, B., & Xue, X. (2026). CAR-VLA: Complexity-Aware and Risk-Adaptive Reasoning for Autonomous Driving. https://omanscience.com/ar/articles/car-vla-complexity-aware-and-risk-adaptive-reasoning-for-autonomous-driving
MLA 9
Chen, Xiaolei, et al. "CAR-VLA: Complexity-Aware and Risk-Adaptive Reasoning for Autonomous Driving." https://omanscience.com/ar/articles/car-vla-complexity-aware-and-risk-adaptive-reasoning-for-autonomous-driving.
شيكاغو (المؤلف–التاريخ)
Chen, Xiaolei, Zhuolin He, Yuxuan Liang, Xu Li, Haotian Chen, Fan Shi, Mengyang Zhao, Wenjuan Meng, Zisheng Chen, Zhihao Zhu, Zhounan Jin, Hengli Wang, Qingfan Wang, Jiamei Liang, Bin Li, and Xiangyang Xue. 2026. "CAR-VLA: Complexity-Aware and Risk-Adaptive Reasoning for Autonomous Driving." https://omanscience.com/ar/articles/car-vla-complexity-aware-and-risk-adaptive-reasoning-for-autonomous-driving.
هارفارد
Chen, X., He, Z., Liang, Y., Li, X., Chen, H., Shi, F., Zhao, M., Meng, W., Chen, Z., Zhu, Z., Jin, Z., Wang, H., Wang, Q., Liang, J., Li, B. and Xue, X. (2026) 'CAR-VLA: Complexity-Aware and Risk-Adaptive Reasoning for Autonomous Driving', Available at: https://omanscience.com/ar/articles/car-vla-complexity-aware-and-risk-adaptive-reasoning-for-autonomous-driving.
فانكوفر
Chen X, He Z, Liang Y, Li X, Chen H, Shi F, et al. CAR-VLA: Complexity-Aware and Risk-Adaptive Reasoning for Autonomous Driving. https://omanscience.com/ar/articles/car-vla-complexity-aware-and-risk-adaptive-reasoning-for-autonomous-driving
IEEE
X. Chen, Z. He, Y. Liang, X. Li, H. Chen, F. Shi, M. Zhao, W. Meng, Z. Chen, Z. Zhu, Z. Jin, H. Wang, Q. Wang, J. Liang, B. Li, and X. Xue, "CAR-VLA: Complexity-Aware and Risk-Adaptive Reasoning for Autonomous Driving," https://omanscience.com/ar/articles/car-vla-complexity-aware-and-risk-adaptive-reasoning-for-autonomous-driving.