A Method for Robust Requirements Definition

Abstract

This paper describes the implementation of the goal-function tree (GFT) for an unmanned aerial vehicle (UAV) in a Systems Modeling Language (SysML) environment which tracks nominal and off-nominal goals for a successful flight for a UAV. The utility of the GFT allows for traceability of the system’s goals and requirements to perform a successful mission scenario. Definition of nominal goals reduces ambiguity about mission success and can be followed more easily by the customer. Furthermore, the goal-function tree can also track and implement corrections if the measurements for success are deviating through the use of off-nominal functions. The GFT’s off-nominal functions create a new set of functions and goals which attempt to mitigate a failing attribute instead of the other methods which only defined failure. This methodology creates a set of procedures for an operator or system when action is needed, thereby identifying more robust requirements as compared to traditional requirements development based only on nominal mission scenarios. The advantage of the SysML environment allows the legible traceability of the overlapping of goals of various structural and behavioral system objects, the execution of functions, and implementation of stereotypes which classifies the interactions between objects.