Model-based support for integration and testing of a multi-disciplinary industrial system

To reduce the integration and test effort for high-tech multi-disciplinary systems, we are developing a method called model-based integration and testing. The method allows integration of models of not yet realized components (e.g. mechanics, electronics, software) with available realizations of other components. The combination of models and realizations is then used for early system analysis by means of validation, verification, and testing. The analysis enables early detection and prevention of problems that would otherwise occur during real integration, resulting in a significant reduction of effort invested in the real integration and testing phases. This paper concerns the application of the method to a relevant industrial case study, involving several model-based analysis techniques that support in clarifying, evaluating, and improving integrated system and component designs, and in evaluating and improving candidate test cases. The models developed allow early prediction of integration and test problems, and the same models help in finding and fixing the root cause of such problems. The case study results contribute to the effort reduction for integration and testing, and encourage further research on the model-based integration and testing method.\u3cp\u3eTo appear in INCOSE Insight magazine\u3c/p\u3e\u3cp\u3eTo obtain a copy of this publication you may contact one of the authors.\u3c/p\u3