Model-based test case generation (MBTCG) is about automatically generating test cases from a computer readable specification. Its advantages are further automation of the test-design process with upsides such as a guaranteed specification coverage and automatic test-case/requirement tracing.
In addition to providing efficient and effective tests, the technique scales to several other contexts.
- Combining MBTCG with binary program verification techniques opens up the ability for highly automated software analysis on machine code level.
- Automated learning of the model through cleverly constructed “probing” test cases opens MBTCG for areas where coming up with a model is difficult. This is interesting in case of verifying large systems of systems or designs that use techniques from the artificial-intelligence research internally.
- Compact generated test suites are useful to review a specification model for unintended corner cases and side effects.
AIT’s MBT tool is available commercially and is being applied to various models of different sizes successfully for functional verification. Current areas of research are test case generation for non-functional requirements and modelling and verification of systems using advanced AI algorithms.
- TCG as a service
- Tool adaptions
- Domain Specific Language development and test case generation support
- Test model development support
- TCG for model development environments