Introducing Requirement Stability Metrics for Test Case Success Prediction in RUAG Space AB

Abstract

Context: In every software development method, requirement gathering and analysis phase plays the most important role. Stability of requirements potentially makes an impact on the success of later phases in a software project, including the success of test cases. Nevertheless, this impact is not well investigated in either theory or industry. Furthermore, the application of software metrics can improve the quality of software and efficiency of software development processes since metrics can help in controlling and making predictions in software development projects. Objective: In this thesis, we intend to introduce requirements stability metrics for test case success prediction in the context of integration and verification unit of RUAG Space AB, Sweden. Method: The research is done by conducting a case study that includes reviewing the related work, defining a set of requirement stability metrics, developing an automated tool for the data collection on a daily basis, and performing empirical evaluations on validity and usefulness of the introduced metrics in an industrial context. Results: The research outcomes present that the proposed requirement stability metrics can be useful for stakeholders after making minor changes in their definitions and the metrics can be applied to integration and verification processes in RUAG Space AB. Conclusions: However, more time is required for data collection to expand the thesis work and to conclude whether the proposed metrics can be used as predictors for test cases successes in RUAG Space AB and other companies. The remaining work can be pursued in future research work.