Data is the source for AD development and validation – Brand new product line for in-vehicle data logging and processing of sensor data that enables data-driven development. Our solution gives you unique performance in terms of bandwidth, support of bus and sensor interfaces, and synchronization.

From recorded data to ground truth information – dSPACE and understand.ai provide a unique and integrated tool chain for automated and high-quality sensor data annotation (camera, lidar, radar) to train and validate neural networks, generate simulation scenarios, and develop functions for autonomous driving.

Reprocessing of real-world data to validate perception and sensor fusion algorithms – Synchronous replay of recorded raw sensor and bus data to validate new software releases and support algorithm validation in the lab with realistic external data using the proven dSPACE real-time technology.

Bringing realism to your simulation – understand.AI and dSPACE provide an AI-based solution for generating complex simulation scenarios from measurement data to ensure a realistic traffic and environment simulation and sufficient test coverage, including critical scenarios from real test drives.

End-to-end simulation ecosystem – Vehicle and traffic simulation featuring ground truth sensors, intelligent driver models, the integration of road networks and environments from HD maps, consistent support of standards, and open interfaces for execution in real time or faster than real time.

Software-in-the loop (SIL) testing on the PC- and in the cloud – simulate and test your software functions or complete V-ECU networks without the “limits of real time” (i.e., faster than real time and/or highly parallel). From classic automotive applications, such as powertrain or brake systems, to e-drive applications and functions for autonomous driving: you can significantly accelerate your software development process by testing and validating virtually.

Lights out and test - Simulating millions of test kilometers over night – Execute the end-to-end simulation tool chain as software-as-a-service. Test and validate all test kilometers with automatically generated scenarios using scalable, virtual test environments with scenario-based test solutions in the cloud.

Connecting simulation and the real world – Validating perception and sensor fusion algorithms by means of physics-based sensor models, real-time simulation, and synchronous sensor raw data injection into high-performance processing units.

Validate radar sensors at the physical level – Closed-loop-testing of radar sensors based on coherent echoes that let you reliably determine the distance, velocity, radar cross section (RCS), and angle of the radar objects.

From photo-realism- to ray-tracing-based sensor models – Virtual reproduction of the entire sensor transmission channel, including propagation aspects such as ray tracing to simulate environments as perceived by camera, radar, and lidar sensors.

High-fidelity lidar simulation and data injection – Testing lidar sensors with injected raw data that is synthetically generated by sensor-realistic simulation, including point-cloud processing. Expand testing capabilities by including real sensors in closed-loop simulations.

Generating real-world environments – Using AI with existing ground truth simulation scenarios with simple 3-D outlines to automatically generate photorealistic images of the traffic environment. The technology demonstrator will simplify testing and validating camera sensors.

Testing LTE and 5G applications in a virtual environment – Integration of sensor data via LTE and 5G networks in open- and closed-loop test setups to provide an end-to-end solution for applications, such as sensor sharing. The solution helps consider the entire chain of effects when validating connected and cooperative autonomous vehicles.

Testing real sensors with simulated radar echoes – Over-the-air stimulation of radar sensors with up to four echoes in a very precise but particularly easy-to-use manner that is suitable for test areas such as end-of-line, homologation, and aftermarket.