
Whether in driver assistance systems, such as pedestrian recognition, in electrical traction control for trains, highly individualized consumer electronics or in the configurative industrial production of machine elements – such highly complex applications require methods and tools with which the quality of the systems can be guaranteed.
In the XIVT project, automated test methods for variant-rich systems are being developed. These test methods are constantly evaluated within the framework of the project and tested in applications of the industry partners of the project in order to confirm the practical usefulness. This ensures that the methods developed contribute to an increased test coverage and more flexible production processes, thereby increasing the overall quality of the products.
Goals
In order to achieve this goal, the first step is to analyze the requirements of various industrial applications areas. Test generation is based on the systemic level: starting from individual single variants, test cases are derived which test the properties on systemic level. In order to use test resources effectively, an application-specific test prioritization is carried out based on existing knowledge about the systems to be tested. A special focus will be placed on natural language processing and machine learning with the aim of simplifying the currently cumbersome process of formulating complicated test cases, which can be only be carried out with professional knowledge, in the long term.