Model-based testing: challenges ahead

In model-based testing, models derived from the informal requirements (or models developed as part of the requirements process) are used to drive the testing; these models are used to generate the tests as well as serve as oracles, and the testing process can be largely automated. This move towards models, tools, and automation holds enormous promise, but it also raises new challenges that, in our experience, must be addressed before we can reap the full benefits.