Integration of agile software development and robust design methodology in optimization of software defect parameters

Software organizations focus on delivery of quality software that meets customer requirements, with zero defects. In agile software development, quality means quality of code wherein the quality is maintained through the use of test-driven development and refactoring. During test driven development of software, large number of defects have been detected and preliminary studies on the causes of these defects revealed the complex nature of the problem. Robust design methodology is one of the important quality improvement techniques used in manufacturing systems focusing on reducing variability in the output. This paper integrates robust design methodology into agile software development in order to capture maximum number of defects at the earliest. The proposed framework is implemented on the defect data from a large telecommunication organization. The software quality is measured in terms of metrics such as number of defects during development of software. Robust design methodology is applied on software during development to determine the optimal parameter values which will influence maximum defects to be captured before the software is released in the market. Three crucial parameters from the dataset are extracted using the analytic hierarchy process. The outcome of the study is the determination of the optimal levels of the input parameters that will maximize the number of defects captured. As agile processes rarely use formal review and inspection processes, the proposed approach will build inherent defect capturing capability in the software development process.