Risk assessment in new software development projects at the front end: a fuzzy logic approach

Abstract

New software development (NSD) has inherent complexity, uncertainty, and risk. Risk analysis and mitigation is perhaps the most critical activity in an NSD project, yet such risk evaluation is often not adequately performed. To reduce the high failure rate of NSD projects, managers require more effective tools for evaluating and managing NSD project risks. Limited by both the nature and timing of NSD, risk evaluation is associated with data, information, and imprecise or ambiguous knowledge. Fuzzy logic is well suited for analysis in this situation. Thus, a fuzzy risk impact rating (FRIR) was developed for determining the total project risk exposure level for an NSD project according to risk attributes associated with the project, such as organizational environment, users, requirements, project complexity, team, and planning and control. The FRIR is composed of attributes’ possible ratings and corresponding severity levels, and is aggregated using fuzzy weighted average. As an illustration, the development of a new electronic toll collection project by a Taiwanese company is evaluated. This evaluation evidences that the fuzzy logic-based risk evaluation model can efficiently aid managers in dealing with ambiguity, imprecision, and complexity in NSD risk evaluation.