dSpace, Aves Reality create virtual 3D worlds for autonomous driving tests
Software provider Aves Reality, a company in the field of geospatial data processing and 3D virtual data worlds, and development tool provider dSpace have entered into a technology partnership. The duo aims to provide autonomous vehicle developers with tools to validate sensors and algorithms during virtual test drives in a variety of realistic scenarios.
For meaningful validations of AD functions and simulations of environmental sensors such as radar and lidar scanners or cameras, extensive three-dimensional environment models are needed that contain realistic locations, materials and semantics. Aves Reality processes satellite imagery with artificial intelligence and advanced video game technology to create 3D content at scale. The Aurelion sensor simulation solution from dSpace uses the provided environments for camera, radar and lidar simulation. This allows algorithms for autonomous driving applications to be validated in simulation in Aurelion using the synthetic environments.
Aves Reality’s 3D environments with a size of several square kilometres enable virtual test drives through entire cities based on existing road networks and real geobase data. The scalable and detailed simulation environments allow development projects to be carried out virtually in a cost-efficient and risk-reduced manner. According to dSpace, results from initial customer projects show that the large-scale digital images of entire city districts offered by Aves Reality add value for vehicle developers.
Aurelion simulation software from dSpace is used across all phases of the development process – for example, software-in-the-loop (SIL) testing, hardware-in-the-loop (HIL) testing or scaled validation in the cloud. Aurelion draws on an extensive library of sensor models. This means that new sensors can be replicated in simulation solutions long before they come onto the market. dSpace has entered into cooperation agreements with the world’s leading sensor manufacturers for this purpose. In addition, it is possible to integrate third-party sensor models.
Accurate environment simulations bring cars faster to the road
Radar target simulator enables separation testing of automotive 4D sensor solutions
Test environment simulates automated valet parking
Cooperation tackles seamless transition from simulation to real testing
Software validates vehicle environment sensors in virtual test drives