Keywords: scenario-based testing, test methodology, toolchain, automatic train operation, automated driving.

Billions of test kilometres are required to validate highly automated driving functions. For economic and organizational reasons, this is not a sustainable solution. To give a remedy a scenario-based approach is pursued in which all possible eventualities are described in scenarios and first tested using simulative tools. Due to the large number of scenarios and the demand for completeness, it is necessary to establish a toolchain for accurate and efficient implementation. Before individual tools can be defined, the theoretical work steps and interfaces of the test methodology must be determined and specified. By reason of the small amount of measurement data available from the real railroad field, a knowledge and process orientated database-supported approach is followed. The generation of scenarios as well as the definition of requirements and the development of simulative and real tests are based on self-recorded measurement data and a well-founded knowledge base. In addition, when defining the chain links, it must also be considered that existing programs for simulating scenarios were primarily developed for the automotive industry. Adapted to existing toolchains from the automotive industry, an iterative test methodology for the evaluation of Automatic Train Operation functions is developed. The centre in the form of databases are the compiled data and insights of different chain links. At the beginning of the toolchain, prior test cases and measurements are performed in order to create a deep understanding of the specific railroad application and to fill the database with qualitative data. Following this, the scenario-based test approach is implemented and the execution of the defined test cases is realized iteratively as a simulation and in the field. In addition, two virtual data manipulation approaches are followed to increase the level of reliability of the system under test. An evaluation of the tests and risk assessment determine the quality of the results and either lead to a contribution of the safety argumentation or extend the database with the gained knowledge.

download the full-text of this paper (PDF, 6 pages, 347 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description