Model based development is used to increase software quality and software development productivity. Models are created at a platform independent level and then transformed to platform specific models by the use of a model compiler using the same principles as when source code is compiled into assembler.

In a model-driven development environment, the engineers should be able to define their test cases at the platform-independent level and be able to estimate or measure the quality of their tests without being concerned of the exact details of the generated source code. The quality of the tests is however, measured based on how well the tests cover the code structure (according to some criterion such as MCDC) and this structure can be very different from the structure of the platform-independent model.

PILOT addresses this problem by defining a methodology for platform-independent testing. The methodology includes means to perform the analysis of code coverage on a platform-independent level as well as tracing the test items between different levels of design. The methods will be implemented in a framework for testing and evaluated at Saab Aeronautics in Linköping.


Project leader

The project in short

Fully name of project:
Platform Independent Level of Testing (PILOT)

2013-09-01 – 2016-03-31

Financed by:
Vinnova NFFP6

Collaborating partners/Partnership:
Saab Aeronautics

Researchers related to the current project

Anders Eriksson, PhD Student

Sten F. Andler, Professor of Informatics

Birgitta Lindström, Senior Lecturer in Informatics

Jeff Offutt, Professor