A Development Procedure for Discovering Uncertain Requirements

Abstract

Uncertain requirements often bring severe negative impacts to system development. Most of the current research focuses on description techniques for uncertainties in requirements and techniques in reducing the uncertainties. However, it is important, in the first place, to identify them at early stages, in order to make a system architecture design sustainable. This paper, instead of reducing uncertainties, goes in the other direction, for the first time that we are aware of, to develop a procedure for discovering as many potential future requirement changes as possible. The results can be fed into the Mission/Requirement Evaluation Space (MRES), a structure that can further be used for trade studies such as architecture optimizations. We will describe the needed tools, and the method using manual and semi-automatic procedures with two use cases, a software and an aircraft example, to demonstrate their usefulness.