Analysis and visualization of sensor raw data is crucial for the development of ADAS / AD systems: it has many facets and is already carried out online during the test drive in the vehicle and offline in the office with the recorded data from the vehicle for further investigation.
The function blocks of AVETO.analyse deliver seamless and performant support beginning from first analysis, via the processing to the test.
For the analysis of measurement data streams, a powerful software framework for the visualization of multi-gigabit sensors must be used. The AVETO visualization software is able to realize this task, online in the vehicle for first checks and later in the laboratory for a deeper analysis. It is capable of processing hundreds of gigabytes of measurement data from multiple sensors. It visualizes camera data, bus data and even 3D point clouds, as is common with Lidar, for example.
To flexibly extend your own visualizations, the framework offers an SDK for the development and integration of customer-specific visualizations.
An integrated software framework in the vehicle, which is used both online and offline in the vehicle, provides a seamless advantage in development.
In the pre-development and prototyping phase, the calculation of e.g. recognition algorithms for sensor data streams in the vehicle is very common. This requires immense performance in the vehicle - and at the same time the hardware must be robust for operation in the vehicle. Therefore, performance and GPU performance of the server class must function in an on-board environment in order to be able to perform AI calculations with sensor raw data currents. b-plus offers suitable hardware platforms for this purpose.