Testing of Safety-critical ADAS Functions with Mobile HiL Platforms
- PDF / 1,501,289 Bytes
- 6 Pages / 595.276 x 790.866 pts Page_size
- 11 Downloads / 166 Views
Testing of Safety-critical ADAS Functions with Mobile HiL Platforms
How can complex traffic scenarios efficiently be verified along the entire test chain for automated driving? Together with ALP.Lab, Virtual Vehicle has created a testing environment that combines real and virtual tests. This includes the mobile HiL and measurement platform Spider, with which the impact of sun position and temperature on camera, radar and lidar sensors can be analyzed.
SELECTION OF SUITABLE TESTS AND TOOLS
Automated Driving (AD) with its numerous Advanced Driver Assistance Systems (ADAS) is accelerating innovation in the automotive and vehicle industry. However, when developing safe ADAS/
36
AD functions, a large number of technical and practical aspects must be taken into account, especially for mixed traffic – which means coexistence of (fully) automated vehicles and cars that are driven by people. From a technical point of view, tests primarily have to provide proof about correct behavior
© Virtual Vehicle
and functional safety of an ADAS/AD function, especially in presence of technical shortcomings. To cover the variant diversity of test scenarios for automated vehicles in the best way possible, the right selection of suitable tests, methods and tools for OEMs, suppliers or sensor manufacturers is a major challenge. Since a onefits-all testing environment is not in sight in the narrower sense, the available options – from collecting real t raffic data to computer simulation to testing the entire vehicle – must be optimally used and aligned. FIGURE 1 presents a simulation environment for objects (pedestrians, vehicle and infrastructure) and ADAS functions; the objects are linked and tested in a cloud,
A U T H OR S
the results passed on to individual simulation nodes in automated driving. COMBINED TESTING COMPETENCE
Dr. techn. Dipl.-Ing. Christian Schwarzl is Head of Dependable Systems Group at Virtual Vehicle Research GmbH in Graz (Austria).
Dipl.-Ing. Gerhard Greiner is CEO of ALP.Lab GmbH in Graz (Austria).
Dr. rer. nat. Paolo Pretto is Principal Researcher for Human Factors & Driving Simulator at Virtual Vehicle Research GmbH in Graz (Austria).
This is where Virtual Vehicle and ALP.Lab come in. The two companies have developed holistic approaches for testing ADAS/AD functions (also AIbased) in a broader context in recent years. In this context, especially computer and test bench simulation is playing an increasingly important role due to the large number of required test scenarios and their practical and economic advantages. Because of the interconnection of ADAS/AD functions within the vehicle, the requirements for integration tests have increased yet again, which makes it necessary to carry out tests as early as possible. The OEMs are also required to provide proof of functional safety for ADAS/AD functions. This applies especially to technical shortcomings of the sensors according to the standard ISO 21448-SOTIF [1], such as the weather-dependent performance. On the one hand, these are difficult to prove, on t
Data Loading...
