A system dynamics based study of software reuse economics

Abstract

Software productivity is a critical issue for Government agencies and the Department of Defense. Satisfying the enormous demand for new software and reducing cost creates pressure to develop new software production techniques. Of these, one of the most promising, strongly supported by Government and DoD agencies, is software reuse. This thesis studies the economics of software reuse using a system dynamics computer model, the Dynamica Reuse Model, which simulates the activities of a software development organization engaged in organization-wide, systematic software reuse. Long-term relationships between reuse rate, productivity, and unit cost are studied by varying consumption cost, production cost, employee turnover rate, and reusable component retirement age. Results suggest long-term steady state relationships may be different from short-term dynamic state relationships. After validation and customization the dynamica Reuse Model can be used to support an organization's cost and schedule software tools. Increasing understanding of the software development process in order to make knowledgeable rather than intuitive predictions about organizational variables related to reuse such as reuse rate, productivity, and unit cost, enable to serve as a management support tool for the complex and costly process of software development.