Abstract
Black box test generation needs a model of the system under test to describe what is to be tested. Testing criteria and test objectives define how it is to be tested. We explain how to represent a system under test via UML models. We derive test objectives from the UML models and use the objectives and the models to generate test cases by mapping them into a problem description for an AI Planner. The planner uses the problem description to generate a test suite that satisfies the UML-derived test objectives.