Factors such as, increasing test coverage, the associated use of resources, and the available time often do not lead to the desired, or even required 100% transparent test coverage. The lack of test coverage increases the risk of overlooking errors, which can have serious consequences.

By using the model-based testing (MBT) a 100% test-coverage is possible. The use of genetic algorithms reduces the number of tests by about 50%. Starting with UML models in EA, test scenarios are then converted by AFRA CASE into test cases.

The advantages are obvious:

Instead of creating many individual test cases, only a few UML models are needed. The high configurability of coverage criteria and other parameters of generation, allow the flexible adaptation to the test requirements (e.g. concerning the number of test cases or length of the test cases).

The generated test cases are platform independent and can be output into any test script format; for example, as manual tests in Excel, Word, or HP / Mercury Test Director or as automatically executable tests for SilkTest or in our own test system AFRA INSPECT.

The automatic generation of test cases and associated test data is used for all phases from the unit test to system test.

Especially the combination of the AFRA CASE and AM|USE (LieberLieber Software) provides additional benefits. AM|USE enables the user for the model execution in the Enterprise Architect directly. On the one hand AM|USE allows simulation of test scenarios generated by the AFRA CASE and on the other hand it can generate profiles (e.g. Happy-Path) for the AFRA CASE for customizing of test generation.

Please find more German information about AFRA CASE here.