发表机构
Rochester Institute of Technology; City University of Hong Kong; New York University（罗切斯特理工学院; 香港城市大学; 纽约大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出SceneFactory-3D，一种GPU批处理的物理基础多智能体驾驶模拟器，通过并行物理反事实评估，研究车辆控制器对道路摩擦和坡度条件的敏感性，发现摩擦降低显著影响安全性和近碰撞频率。
AI 中文摘要
可扩展的驾驶模拟器通常使用预设的行为或运动学规则来执行车辆指令，忽略了轮胎-路面界面的物理特性，从而限制了它们捕捉不利道路和环境条件如何改变车辆执行并在交通中传播的能力。为解决这一局限性，我们提出了SceneFactory-3D，一个基于GPU批处理、物理基础的多智能体驾驶模拟器。车辆通过在每个车轮接触点评估的悬架和摩擦力限制力来执行加速和转向指令。空间变化的摩擦力、每个世界的3D高度场、重力和刚性接触一致地控制车轮运动和底盘碰撞。每个世界的地形隔离和GPU批处理使SceneFactory-3D能够并行运行匹配的物理反事实：交通场景设置和车辆控制器保持固定，仅道路条件变化，从而可以在平行世界中评估由此产生的闭环效应。为展示SceneFactory-3D支持的反事实评估的优势，我们对车辆控制器对道路条件的敏感性进行了实证研究。我们在每种条件下研究了三个学习策略族在1,024个匹配的12车世界中，以及两个经典规划器在共享的32世界子集上，跨越21种摩擦和坡度条件。当摩擦从1.0降至0.18时，清除工作区安全的车辆比例在学习策略中下降了6至90个百分点（经典规划器为18-19个百分点），并且每个学习策略的近碰撞情况变得更加频繁。代码：此https URL
英文摘要
Scalable driving simulators typically execute vehicle commands using prescribed behavioral or kinematic rules, overlooking the physics of tire-road interfaces, thereby limiting their ability to capture how adverse road and environmental conditions alter vehicle execution and propagate through traffic. To address this limitation, we present SceneFactory-3D, a GPU-batched, physics-grounded multi-agent driving simulator. Vehicles execute acceleration and steering commands via suspension- and friction-limited forces evaluated at each wheel-contact point. Spatially varying friction, per-world 3D heightfields, gravity, and rigid contact consistently govern wheel motion and chassis collisions. Per-world terrain isolation and GPU batching enable SceneFactory-3D to run matched physical counterfactuals in parallel: traffic scenario setup and vehicle controllers remain fixed while only the road condition changes, enabling the resulting closed-loop effects to be evaluated across parallel worlds. To demonstrate the advantage of the SceneFactory-3D-enabled counterfactual evaluation, we conduct an empirical study on vehicle controllers' sensitivity to road conditions. We study three learned-policy families on 1,024 matched 12-vehicle worlds per condition, and two classical planners on a shared 32-world subset, across 21 friction and grade conditions. When friction drops from 1.0 to 0.18, the share of vehicles that clear the work zone safely falls by 6 to 90 percentage points across learned policies (18-19 for classical planners), and near-collision situations become more frequent for every learned policy. Code: https://github.com/SmallWorldLab/SceneFactory_3D
