Development solutions for autonomous driving
Tools, software and services for electronic systems
The requirements/use cases for autonomous driving are very complex.
Accordingly, today’s software development tools must be capable of supporting the complexity associated with the increasing number of cameras, sensors and image processing requirements and their networking. ETAS development solutions help to master these development challenges faster and more efficiently.
For example, virtualization allows development steps to be performed earlier and in parallel, hardware and test vehicles can be saved, test sequences can be accelerated, and safety-critical tests can be performed safely – a major advantage, especially when safeguarding highly automated driving functions.
An integral part of system development are various activities for validation and optimization such as functional validation, safety tests and calibration. These are mainly performed on real, hardware-based systems. This can be a HiL-System, a test bench or a test vehicle. Using model-based techniques, these activities can be executed in virtual setups parallel to the system development and preferably, integrated into the continuous development and integration CI-CD pipeline. To enable these activities in a virtual setup ("frontloading") there are many prerequisites:
- Virtual ECUs (vECU) can be created before or in parallel with the real ECU-HEX file.
- Complex systems consisting of DVE models, virtual ECUs and virtual networks can be simulated.
- Parameterized models are available to represent the DVE fit as a replacement for real setups, including DoE approach.
- The simulation of heterogeneous virtual ECUs, both in closed and open control loops, is possible.
- Calibration and network interfaces are provided
- Accurate rest bus simulations are available
- Open and standardized interfaces for the integration of 3rd party tools and models are available.
- The technology exists to enable vast amounts of parallel simulations, including cloud-based approaches.
ETAS offers virtualization solutions that cover all these use cases. Visit our booth and learn more!
GETK-Px for fast data acquisition from ADAS ECUs
A measuring system with uniquely high performance to meet the requirements for data acquisition from HAD control units.
Depending on driving habits, a single autonomous vehicle generates over 300 TB of ECU and sensor data per year. The recording of this data is particularly important during the development phase of the control unit.
Measurement tools that can record 0.5 GByte/s at the ECU are currently still rare. In the ADAS environment and especially with regard to the development of highly autonomous vehicles, data rates of several gigabit/s are in demand.
The new, high-performance GETK-Px interface series from ETAS fills this gap precisely.
The GETK-Px offers the following physical interfaces to the ECU:
- AURORA & DAP to the ECU-µC
- PCIe Gen3 to ECU-µP
The captured data is packaged into Ethernet frames and forwarded via 2x 10 Gigabit Ethernet interfaces from the GETK-Px to a logger system.
The total net data rate is up to 1.5 GByte/s per GETK-Px!
The special feature: The measurement software does not run on the ECU, but exclusively on the ETAS GETK-Px data acquisition hardware to reduce the load on the ECU CPU.
As a result, more computing power is available on the ECU for complicated calculations.