rFpro has announced that Advanced Micro Devices (AMD), a provider of high-performance and adaptive computing solutions, has adopted its AV elevate simulation platform to accelerate the development of new automated driving technologies. With the platform, AMD intends to utilize its simulation capabilities to tune, test, and refine its own automated parking systems on a variety of parking scenarios while more broadly reducing reliance on physical testing.
This announcement builds on a demonstration of the parking solution carried out at CES 2025, where AMD used rFpro’s high-fidelity simulations to showcase how its parking assistance technology handled different virtual scenarios. The demonstration itself was focused on highlighting the technology’s ability to handle complex parking scenarios such as tight spaces, challenging maneuvers, and various lighting conditions.
AV elevate offers a fully integrated simulation solution that enables customers to tune sensors, train their perception and control algorithms, and conduct testing across the full AV technology stack. In carrying out this work, it provides both closed-loop perception testing and can create engineering-grade synthetic training data. The platform more broadly integrates high-fidelity sensor models for all major types of sensors commonly found on autonomous vehicles, including cameras, LiDAR, and radar, and includes a broad library of standard sensor models alongside digital twins of commercially available sensors.
rFpros’ physically modelled virtual environments and ray tracing rendering technology sit at the core of AV elevate. Together, they allow the elements within the scene to be physically modelled with realistic material characteristics and a road surface model accurate to a millimeter, enabling the production of more realistic training data.
Working alongside rFpro’s environments and ray tracing technologies is the company’s library of real-world digital twins, designed to help facilitate more accurate, and diverse, virtual proving grounds. Through these digital twins, AMD will be able to access a broad range of locations to simulate complex, geo-specific, scenarios. Among these locations are typical underground, city and rural parking garages, as well extensive real-world public roads.