Demonstrated for the first time at CES 2022, the Radar Scene Emulator from Keysight Technologies Inc. represents another step ahead for autonomous vehicles: Its full-scene emulation allows lab testing of driving scenarios as complex as the real world.
Keysight said the new device combines hundreds of miniature radio frequency (RF) front ends into a scalable emulation screen representing up to 512 objects and distances as close as 1.5 meters. It also employs patented technology that shifts from an approach centered on object detection via target simulation to traffic scene emulation — thus offering several key benefits for OEMs:
- Big-picture view: Full-scene rendering that emulates targets both near and far across a wide, continuous field of view (FOV) eliminates radar vision gaps. It also enables better training of algorithms to detect and differentiate multiple objects in dense, complex scenes. The result is that autonomous vehicle decisions can be made based on a complete picture — not just what the test equipment sees.
- Real-world complexity: Whereas testing radar sensors against a limited number of targets provides an incomplete view of driving scenarios and masks the complexity of the real-world, the new device enables driving scenes in the lab to be emulated with variations of traffic density, speed, distance and total number of targets.
- Learning acceleration: Complex scenes that today can only be tested on the road are supported by the emulator’s real-world environment, allowing OEMs to accelerate algorithm learning for advanced driver assistance systems (ADAS) and autonomous driving (AD) capabilities. Testing scenarios can include stationary or in-motion objects and varying environmental characteristics, while eliminating inefficiencies from manual or robotic automation.
- Improved scene resolution: Testing the ability to distinguish between obstacles on the road — a key characteristic for Levels 4 and 5 autonomation as defined by the Society of Automotive Engineers (SAE) — is supported by point clouds, or multiple reflections per object.
Keysight’s Radar Scene Emulator is part of the company’s Autonomous Drive Emulation (ADE) open platform, created through a multi-year collaboration between Keysight, IPG Automotive and Nordsys. The platform can be integrated by automotive OEMs with commercial 3D modeling, hardware-in-the-loop (HIL) systems and existing test-and-simulation environments.
As Thomas Goetzl, vice president and general manager for Keysight’s Automotive & Energy Solutions business unit, said: “The vision of fully autonomous vehicles is rapidly approaching, and we’re thrilled to be accelerating this vision into a reality.”