Optimal release policy under fuzzy environment

Computer systems forms the fundamental part of the modern information society. Building software possessing highest degree of quality, based on latest technology and offering it at an economical price is a challenging job. Many decision-making (DM) stages arise during the process and the decisions based on system information, data and quantitative analysis are preferred. An important decision that the management must take is to decide when to stop testing and release the software so as to make a timely release and obtain maximum returns on the organizational objectives. The existing research in software release time decision problem using classical optimization methods require well-defined criterion and activity constant coefficients, resource, requirement and structural constants together with well-defined inequalities. However in the practical applications computation of these constants depend on many non-deterministic factors bringing uncertainty in their exact computation. If we use the values calculated assuming deterministic system environment, the solutions of the model may not reflect the real picture and may lead to incorrect DM and huge losses. Fuzzy models offer the opportunity to define subjective imaginations of the decision makers/computations of the model constants more precisely. The importance of formulating optimization models under fuzzy environment forms the main motivation of our study. In this paper, we have formulated software release time decision problems under fuzzy environment and have discussed the solution methodology based on fuzzy set theory and fuzzy optimization through numerical illustration. © The Society for Reliability Engineering.