Abstract
Embedded software is getting larger and more complex. In addition, when it is developed, high quality, short of time to market and lower costs are required. In order to achieve these severe nonfunctional requirements, we need to understand it quantitatively and improve the process for its production. We show that COSMIC-FFP is useful for quantifying embedded software. COSMIC-FFP, which is an extension of function point analysis in order to apply the method to real time software, is a type of a functional size measurement. The technique extended for real time software is also useful for embedded software. In this paper, we mention the characteristic of embedded software and issues when quantifying embedded software, and how COSMIC-FFP can solve them. And we illustrate some trials use of COSMIC-FFP done by ourselves in order that COSMIC-FFP is useful for embedded software.