Testing is a paramount quality assurance activity in every software development project, especially for embedded, safety critical systems. During the test process, a lot of effort is put into the generation of test cases. The presented OSTAG project aimed at developing methods and techniques to automate the software test case generation for black-box testing. The proposed approach was based on the creation of a software requirements ontology and the application of inference rules on the ontology to derive test cases. The ontology represents knowledge of the requirements, the software system and the corresponding application domain while the inference rules formalize knowledge from documents and experienced testers in the domain of test planning and test case generation. A software prototype of the approach was implemented and one of the industrial project partners evaluated the results. An alternative method for generating test cases, based on genetic algorithms, was also explored.