汽车前沿技术 | rFpro推出仿真解决方案AV elevate IN CABIN

来源：市场资讯

据报道，仿真软件公司rFpro推出 AV elevate IN CABIN。这款附加软件包使汽车原始设备制造商（OEM）、一级供应商（Tier 1）和传感器开发商能够在仿真环境中对驾驶员和乘员监控系统进行调优、训练和测试，从而提高车内测试的速度和一致性，并可在任何物理原型车制造之前就开始进行测试。

根据Euro NCAP 2026年的新规，驾驶员和乘员监控系统在车辆安全评级中的权重显著提高。OEM及其供应商面临的挑战是，这些系统必须在各种实际路况下可靠运行，而物理测试耗时、成本高昂且范围有限。

rFpro技术总监Matt Daley表示：“AV elevate IN CABIN是一款物理精度极高的工程级仿真环境，能够在实物制造出来之前进行数千次测试。”

rFpro的全新AV elevate IN CABIN解决了车内仿真的三个核心问题：虚拟座舱环境的逼真度、乘员行为的控制以及传感器感知的物理精度。

车内环境逼真度

rFpro通过添加人眼和摄像头不可见但雷达可探测到的子结构，提升了车辆内饰模型的逼真度。例如，座椅内的金属框架现在经过建模，可以确保基于雷达的传感系统能够接收到真实的反射信号。

rFpro的车内物体库也得到了扩展，新增了车内常见的物品，例如背包、笔记本电脑、儿童座椅、宠物和其他个人物品。

皮肤仿真也得到了改进，采用了更精细的面部模型，以支持通过面部特征评估驾驶员状态的系统。

虚拟环境控制

rFpro人体模型中的高密度骨骼绑定系统能够精细控制面部和四肢的运动，从而仿真驾驶员监控系统必须检测的宏观和微观表情。这包括“猫头鹰式动作”（驾驶员转头）和“蜥蜴式动作”（仅眼球离开路面）。

Daley表示：“我们甚至考虑到了车窗的开关等细节。对于外部传感器仿真来说，这并非需要考虑的问题，但它会从根本上改变光线和雷达能量在车厢内的传播方式。对于想要开发在现实世界中可靠运行的车内系统的开发者而言，这些是必须考虑的变量。”

传感器建模

大多数车内系统都采用雷达传感器、可见光摄像头和红外（IR）摄像头的组合，这些摄像头利用各自的IR能量源“观察”周围环境。RFpro的新型IR摄像头传感器集成方案能够精确模拟摄像头发射的能量及其与车内所有表面的相互作用。所有内饰材料，包括驾驶员皮肤、座椅和车窗，都已被赋予特定的IR反射率和雷达特性。这些材料特性已与Sim4CamSens研究项目下进行的实验室测量结果进行关联，该项目由英国国家物理实验室（National Physical Laboratory）和化合物半导体应用弹射器（Compound Semiconductor Applications Catapult）参与测试。例如，皮肤反射率的特性已细化到鼻子、下巴和脸颊之间的差异，以提高IR摄像头数据的真实性。

Euro NCAP将提供一套含预配置场景的测试包，涵盖所有基本的驾驶员和乘员监测评估。rFpro是目前唯一支持在驾驶员在环（driver-in-the-loop）、硬件在环（hardware-in-the-loop）和软件在环（software-in-the-loop）环境下进行外部和内部传感器开发的仿真平台，这意味着继用于开发ADAS和自动驾驶系统后，该平台现在也能用于车内传感。

Daley补充说道：“展望未来，了解车内人员及其行为对于自动驾驶车辆的运行至关重要，这不仅关乎安全，也关乎乘客体验。如果知道乘员的位置，就可以优化从安全气囊展开到降噪音频等所有功能。我们期待与传感器开发商和OEM合作，进一步推动这项技术的发展。”

Simulation software specialist, rFpro, has launched AV elevate IN CABIN. The add on package enables automotive OEMs, Tier 1’s and sensor developers to tune, train and test driver and occupant monitoring systems in simulation, improving the speed and consistency of in-cabin testing, and starting long before any physical prototype exists.

Under Euro NCAP’s new 2026 protocols, the weighting of driver and occupant monitoring systems on a vehicle’s safety rating has significantly increased. For example, Driver Monitoring Systems has increased from a maximum of two points to 25 and child presence detection has increased from four points to five. At the same time, Advanced Driver Distraction Warning (ADDW) becomes mandatory for all new vehicles from July 2026 under the EU’s General Safety Regulation. The challenge for OEMs and their suppliers is that these systems must work reliably across a vast range of real-world conditions and testing them physically is slow, expensive and limited in scope.

“Euro NCAP’s 2026 scoring changes make in-cabin monitoring one of the most consequential areas of vehicle safety development. When you combine that with ADDW becoming mandatory, it is clear that the demand to develop and validate these systems faster and earlier in the programme will only grow,” said Matt Daley, Technical Director at rFpro. “We have taken our proven exterior simulation solution and applied the same techniques to the interior. AV elevate IN CABIN is a physically accurate, engineering-gradesimulation environment enabling thousands of tests to be conducted before anything physical has even been built.”

rFpro’s new AV elevate IN CABIN addresses three core areas of in-cabin simulation: the fidelity of the virtual cabin environment, the control of occupant behaviour within it and the physical accuracy of sensor perception.

Interior fidelity

rFpro has improved the fidelity of its vehicle interior models by adding sub-structures that are invisible to the human eye and to cameras but are detected by radar. The metal framework within seats, for example, is now modelled to ensure radar-based sensing systems encounter realistic returns.

rFpro’s in-cabin object library has also been expanded to include items commonly found inside vehicles, including rucksacks, laptops, child seats, pets and other personal belongings.

Skin simulation has also been improved, with higher-detail facial models to support systems that assess driver state from facial features.

Controlling the virtual environment

A high-density bone rig within rFpro’s human models enables fine control of facial and limb movements, supporting simulation of the macro and micro expressions that driver monitoring systems must detect. This includes ‘owl movement’, where the driver turns their head, and ‘lizard movement’, where only the eyes shift away from the road.

“We have even considered details like opening and closing vehicle windows,” continued Daley. “That is not something you need to think about for external sensor simulation, but it fundamentally changes how light and radar energy travel through the cabin. If you are developing an in-cabin system that needs to work reliably in the real world, these are the kinds of variables you have to account for.”

Sensor modelling

Most in-cabin systems use a mixture of radar sensors, visible light cameras and IR cameras, that ‘see’ using their own IR energy source. rFpro’s new IR camera sensor integration accurately models the energy emitted by the camera and how it interacts with every cabin surface. Specific IR reflectivity and radar properties have been assigned to all interior materials, including the driver’s skin, seats and windows. These material characteristics have been correlated with laboratory measurements conducted under the Sim4CamSens research programme, with the National Physical Laboratory and Compound Semiconductor Applications Catapult involved in the testing. Skin reflectivity, for example, has been characterised down to the difference between a nose, chin and cheek to improve the realism of IR camera data.

A Euro NCAP test package with pre-configured scenarios covering all base driver and occupant monitoring assessments will be made available. rFpro is now the only simulation platform supporting both external and internal sensor development within driver-in-the-loop, hardware-in-the-loop and software-in-the-loop environments, meaning the same platform already used to develop ADAS and autonomous driving systems can now cover in-cabin sensing too.

“Looking further ahead, understanding who is in the cabin and what they are doing is essential for autonomous vehicle operations, not just for safety, but for the passenger experience. If you know where occupants are, you can optimise everything from airbag deployment to noise-cancelling audio,” added Daley. “We are looking to partner with sensor developers and OEMs to help drive the direction of this capability further.”

作者：autotechpr

来源：https://www.autotechpr.com