Subproject 3: Testing
The subproject “Testing” prepares methods and tools for carrying out tests in the laboratory, at the testing site as well as in real traffic situations and then demonstrates these in a practical manner.
The highly-automated driving functions (HAF) require a safeguarding concept from various coordinated tools. Against the background of an adequate testing coverage, the subproject represents the useful interplay between traffic scenarios, laboratory tests and simulation, site testing as well as field safeguarding. The goal is to answer essential questions, regarding what must be tested for each examining scenario.
A further goal is the central databank of test specifications, for the test case maintenance and management. This serves, in particular, as reference, with the goal of a comprehensive description of the testing scope for the demonstration of the adequate control of the highly-automated driving tasks through HAF. For the efficiency and process reliability of the transfer of results to future users, an electronic transmission is intended, in an adequate, automatically readable format.
In this way, the scenario classes of the traffic environment will be assessed, in accordance to the probability of occurrence and hazard potential. The verification that these situations are controlled in adequate manner, should be carried out, with the help of the developed test methods and procedures. Reference tests with a high degree of novelty and/or particular relevance for the safety verification will be implemented and evaluated, exemplary, in the simulation, as well as at the testing site. The validation of the results is then carried out in the field trial, in which new scenarios are recorded and imported into the test specifications databank, for further test runs.
Simulation is an essential building block for the proof of safety regarding a HAF function. Thus, for example, the highway chauffeur includes significantly more possible driving situations, in comparison to the previously partially-automated driver assistance systems. Including the possible environmental conditions for a HAF vehicle (for example, the variation of the static traffic environment, the other traffic participants and their driving behavior, of the weather), an extremely large testing space will result. Therefore, the goal is to master the same level of safeguarding, with the help of the new virtual testing methods.
The simulated testing situations should be verified and, thus, the simulation method be validated. Given such matters, the situations from the test site should be reproduced, within the limits of controllability. Therefore, it involves the hard-to-control and complex traffic situations, with several involved vehicles. At the testing site, measurements must be carried out, which cannot be put into practice by simulations alone.