Simulation software specialist rFpro is playing a key role in Sim4CamSens2, a new collaborative research project designed to advance the development and validation of automotive sensor systems through simulation. The initiative builds on the success of the original Sim4CamSens project and extends the focus from exterior to also include interior-facing sensor systems, addressing the growing importance of in-cabin monitoring for safety and autonomy.
Led by Claytex, the project brings together partners including Oxford RF, Syslek, the Queen Mary University of London, National Physical Laboratory (NPL), Compound Semiconductor Applications (CSA) Catapult and AESIN. rFpro’s contribution will further advance simulation verification processes, and research how synthetic data can be used for assessing the performance of real-world deployed perception systems and enhancing its simulation platform to model and test in-cabin sensor systems with greater fidelity than previously possible.
The first Sim4CamSens project aimed to improve the realism of external-facing sensor simulation. rFpro upgraded its platform to more accurately replicate real-world light attenuation, reflections and environmental conditions, such as fog. These simulations were correlated with physical sensor data across a wide range of lighting and weather conditions.
Critically, improvements to rFpro’s rendering engine made synthetic data appear more realistic to AI systems, increasing the confidence that perception system developers can place in simulated data.
“With the first phase of the project, we demonstrated that high-fidelity simulation can closely match real-world sensor behaviour across diverse lighting and weather conditions,” said Matt Daley, technical director at rFpro. “That correlation is vital to encourage the industry to embrace synthetic data as a legitimate tool for tuning, training and testing their systems. This will massively accelerate development and reduce the need for relying on physical data collection and prolonged physical testing.”
Sim4CamSens2 continues this trajectory by tackling two new challenges. Firstly, the project will develop more efficient sensor characterization methods. Rather than requiring extensive field testing over months, the initiative aims to establish minimal laboratory testing protocols that provide sufficient data for accurate simulation correlation. This will significantly reduce the time and cost of developing new sensor integrations into the simulation.
Secondly, it will examine how perception systems perform in degraded conditions, such as a scratched or dirty lens or various levels of fog, in both real and synthetic data. This will help define the conditions that most impact system performance, enabling engineers to focus development efforts where they matter most.
The project also marks rFpro’s first major push into interior sensor simulation. With in-cabin sensing playing a growing role in both safety and autonomy. Sensor systems are needed for a diverse range of tasks, including to measure occupant size and posture for restraint system optimization, monitor driver attention and, for autonomous vehicles, verify passenger readiness before departure. The rapidly evolving in-cabin experience market is also using sensing technology to deliver the next generation of features that will differentiate the automotive products of tomorrow.
rFpro will adapt its simulation tools to model human movement and skin appearance, develop higher detailed digital twins of vehicle interiors, and create in-cabin-specific edge case scenarios for testing. It will also develop new sensor type integrations for infrared cameras, introducing new technologies to the rFpro platform.
These developments will support AV Elevate, rFpro’s award-winning simulation platform for ADAS and autonomous vehicle development, extending its capabilities to include interior sensor tuning, training and testing, alongside existing exterior sensor simulation.
“A lot of our existing technologies can be applied to interior-facing sensors with some adaptation,” said Daley. “By covering both exterior and interior sensing, we’re providing the industry with the most advanced single simulation platform to accelerate the safe deployment of intelligent and autonomous vehicles.”
Several Tier 1 automotive suppliers are already collaborating with rFpro on this technology development, establishing a clear commercialization pathway and demonstrating existing market demand for enhanced interior sensor simulation capabilities.
SIM4CAMSens 2 is funded through the £150m CAM Pathfinder program announced in the UK government’s Advanced Manufacturing Sector Plan. The program is funded by the Centre for Connected and Autonomous Vehicles (CCAV), delivered in partnership with Innovate UK and Zenzic.
The CAM Pathfinder 1 – Enhancements competition was launched in December 2024 to target early commercial self-driving vehicle opportunities and support the UK supply chain to grow and fill technology gaps necessary for their deployment, and is part of the government’s vision for self-driving vehicles (Connected and automated mobility 2025: realising the benefits of self-driving vehicles).
Mark Cracknell, program director at Zenzic, said, “We’re delighted to announce the SIM4CAMSens2 project as one of the eight exciting CAM Pathfinder – Enhancements projects taking place across the UK. The deployment of connected and automated mobility solutions holds incredible promise – enhancing accessibility, reducing emissions, and fostering a transport network that is both reliable and inclusive.
“The SIM4CAMSens2 project will serve as a fantastic showcase of these benefits. We are looking forward to working with the project consortia as they further develop their work, showcase their technology, and help to cement CAM as an industry of the future.”
In related news, Gatik has launched Arena, its next-generation simulation platform designed to accelerate the development and validation of autonomous vehicle systems. The platform has been developed in collaboration with Nvidia and integrates Nvidia Cosmos World Foundation Models (WFMs). Read the full story