NXP India develops semiconductors for a wide range of applications, including automotive radar sensors. In an interview, Suresh Vadde, BL-AAES System SW Validation group Manager (ADAS and Infotainment) at NXP in Bangalore explains the role played by dSPACE solutions.
Mr. Vadde, what are NXP's main areas of activity in India?
Suresh Vadde: NXP in India, at the Bangalore site, is a tier 2 company that develops semiconductor products for a wide range of consumer and industrial applications. This broad spectrum also includes chips for radar applications in the automotive field, for example the SAF85xx RFCMOS automotive radar System-on-Chip (SoC) (Figure 1), which enables autonomous driving or driver assistance systems. Buzzwords in this context include automatic emergency braking, adaptive cruise control, lane departure warning, and cross-traffic warning, to name but a few. In addition, we also create reference designs and reference tools for our chips. These can then be used by the respective tier 1 companies in the actual radar sensor development. For this, we offer our customers evaluation boards, antenna reference designs, and sample software.
Which role do dSPACE solutions play in this?
Suresh Vadde: We can use them to test the radar thoroughly and comprehensively on functional level in the laboratory. This significantly reduces the number of real test drives on the road and therefore also the costs.
The advantage in the laboratory is the exact reproducibility and definability of the scenario down to the last details – something that is difficult to achieve on the road due to external influences. The dSPACE equipment also offers the possibility of automation for functional tests around the clock and access from anywhere in the world thanks to remote access. And because this can all be done with very early functional prototypes, we save a lot of time, because the earlier you can detect a possible error, the easier it is to correct it.
And which dSPACE products do you actually work with?
Suresh Vadde: On the hardware side, the following should be mentioned in particular: A radar test bench with various DARTS for simulating the radar targets and a SCALEXIO HIL simulator for controlling and automating the test bench in real time and creating an interface to the ECU. With this setup, we realize a reliable multi-target, multi-angle simulation for testing complex ADAS/AD scenarios.
On the software side, we use AURELION for realistic visualization and Automotive Simulation Models (ASM) for the simulation of traffic scenarios. Test automation for round-the-clock tests takes place on the basis of dSPACE AutomationDesk.
ConfigurationDesk, ControlDesk, and ModelDesk are also used as additional software packages, including for overall configuration and experiment control.
Keyword DARTS: Can you explain these systems in more detail?
Suresh Vadde: DARTS stands for dSPACE Automotive Radar Test Systems. With DARTS, radar sensors can be tested in the laboratory in clearly defined, reproducible scenarios. DARTS simulate object reflections – so-called radar targets – in real time, which occur in road traffic at different distances, speeds, and sizes. Specifically, we use the DARTS 9030-MS and the new DARTS 9040-GT, which is capable of simulating two targets with a bandwidth of 5 GHz. The possibility of modular use of the DARTS is also a great advantage, as we can also install them in our other test systems to simulate dynamic targets using DARTS standalone. The Compact 3D radar test bench can even remain in use, for example, with the DARTS 9030-MS in a different chamber and the DARTS 9040-GT remaining in the Compact 3D radar test bench. This allows the system to be flexibly adapted to various test scenarios.
You use a wide range of dSPACE hardware and software products – what has been your experience so far?
Suresh Vadde: In short: Because all the hardware and software used for functional-level testing in this setup comes from dSPACE, i.e., from a single source, all the components work together perfectly. We therefore have none of the interface problems that often occur in development environments with products from different manufacturers.
What particular challenges did you have to overcome when setting up your development environment and how did you master them all?
Suresh Vadde: One of the main challenges is the usability of different ECUs in the test environment. Depending on the radar control unit, this requires the adaptation of suitable software interfaces (restbus simulation) and the design of new hardware interfaces for the device under test (DUT). Because the DUT can vary greatly, the range of possible test cases is also very wide.
The dSPACE system offers the necessary flexibility, as the DARTS are agnostic with respect to the radar sensor signal. Only the polarization needs to be taken into account, but this is easy to adjust. The mounting plates for the DUT can also be easily designed according to the dimensions of the hardware. The various test cases for the respective DUT can be created with the help of dSPACE software functionalities. This enables functional tests of the radar firmware directly on the intended hardware due to the real radar targets generated by the test system. Another important aspect is the possibility of remote operation of the dSPACE system, which means that it can also be used by developers at other locations around the world.
What were the main reasons for you to choose the dSPACE test environment?
Suresh Vadde: The decisive factor in selecting the dSPACE solution was that it combines the reliability of a proven, industry-tested system with a high degree of flexibility. The radar test bench from dSPACE enables realistic real-time simulation of radar targets in a variety of traffic scenarios for black-box testing of radar-based ADAS/AD functions using over-the-air stimulation in a multi-target, multi-angle simulation. The option for automated round-the-clock testing and remote access is particularly useful. A further plus point is that dSPACE in India has a branch in Bangalore and other offices at additional locations, which simplifies engineering support on site.
Profile: Radar Test Bench – Compact 3D
The "Radar Test Bench – Compact 3D" enables realistic testing of radar-based vehicle functions and verification testing of radar sensors and their components.
- Over-the-air stimulation of the radar sensor
- Reliable multi-target angle simulation for up to 20 radar objects in real time
- Test of the entire ADAS/AD chain of effects
- Support for 24-, 60-, and 77-GHz radar devices
- Anechoic absorber chamber avoids unwanted reflections
What steps are you planning next?
Suresh Vadde: The automation of the radar functional tests, which we will realize with the help of dSPACE AutomationDesk or via test scripts. This enables automatic round-the-clock testing, which of course saves a lot of time and money. And because the radar test bench has proven itself so far, we want to integrate it even more extensively into the NXP test infrastructure. We are planning to adapt it to other radar control units from NXP via standardized software and hardware interfaces.
With the Compact 3D radar test bench from dSPACE, we can effectively realize multi-target, multi-angle simulations for reproducible tests of ADAS/AD functions in complex scenarios, even remotely and automatically around the clock.
About the interviewed:
Suresh Vadde
BL-AAES System SW Validation Group Manager (ADAS and Infotainment), NXP India
