Statechum is a general-purpose tool for state machine inference and test generation. It comes with Erlang front-end to make it possible to "learn" state-transition diagrams from Erlang software, using both exported functions and OTP calls.

Given a series of traces (such as those obtained from logs), Statechum attempts to infer a state machine compatible with those traces. This amounts to finding out where loops are but is heuristical in nature because there is usually not enough traces to provide a solid foundation for such decisions. This is where testing is helpful: every time a learner makes a hypothesis, it generates a series of tests to confirm it. Where a hypothesis proves to be wrong, the learner takes this into account and generates a new one. The generated tests complement the original traces. At the end of the learning process, one ends up with a state machine and a series of tests.

In the tutorial I'll present the tool using an example and will show both its learning capability and the support for comparison of state machines where a "diff" between two models is computed.