A fully equipped vehicle with automated driving functions can contain more than 20 different sensors. Using simulation to test these vehicles requires that the simulated sensors perceive the virtual environment with the same accuracy as the sensors in the real world.
Therefore, our simulation environment supports the installation of further sensor types, including high-resolution camera and lidar simulation, in addition to the synthetically generated ground truth. With our wide range of 3D assets we provide a virtual environment in that sensor developers can test, if their simulated sensors see what they should see: what is the object relevant for me and which material is it made of.
For a fast implementation, the simulation environment for the sensor simulation can be created procedurally based on available assets from our database. However, if a one-to-one copy of a specific environment is required, our team will be happy to help you create accurate reproductions.
We attach great importance to optimizing the simulation environment for the sensor simulation, e.g. by storing meta information needed to detect the reflectivity of different materials. At last it can safe lifes if an autonomous driving car tries to drive through a pile of leaves or stones, for example.