Search

The Defense Advanced Research Projects Agency (DARPA) has recently begun an effort to further refine its Fractionated Spacecraft vision. This vision (called the "F6" program) seeks to explore distributed spacecraft architectures capable of performing complex functions similar to current monolithic spacecrafts. However, the ability of fractionated systems to work separately would provide significant benefits not achievable by monolithic systems. The question of interest to this study was "Can these benefits be quantified and evaluated in a meaningful fashion in order to compare them to monolithic architectures?" Rather than narrowing the approach to one or two concepts, the strategy was to expand the decision space in order to generate and evaluate a variety of alternatives under different scenarios and design choices. The stimuli were specifically selected to capture many issues not included in such studies such as technology obsolescence, funding profile changes, equipment failure scenarios, and even changes to mission objectives during the mission. Each stimulus has a variety of decision options or responses which a manager can exercise including delay, cancellation, acceleration of a development, or even new developments. It is argued that the very nature of fractionation will result in systems far more agile to external (and internal) stimuli, but this has never been shown explicitly. This study explores the use of Real Options in conjunction with traditional Net Present Value techniques to quantify the desirability of alternative candidate designs in terms of their adaptability and survivability. Our approach to evaluating the business case for each use case was to explicitly model both the implementation and operation phases for the life cycle of a fractionated cluster. The models were integrated into Phoenix Integration's ModelCenter® framework and used to generate Implementation Value Metrics (IVMs) and Operational Value Metrics (OVMs) associated with the use cases under evaluation. A number of models were developed to support these evaluations using Excel, Matlab, and Arena® (a commercial discrete-event simulation software product). The Model-Based Systems Engineering (MBSE) approach used in this study yielded transformations between high-level frameworks (OWL and SysML) and the models used to generate the data we needed for the evaluation. This rule-based transformation process will be critical for exploring the large decision trade spaces in a reasonable amount of time. This paper presents the results of the business case analyses; the underlying OVM's and IVM's, the approaches used to model the problem, explore the tradespace, refine the exploration, and generate results. We also discuss possible directions for future work. Specifically, it should be noted that while the approach described herein is being developed for, and funded by, DARPA, it has general applicability to a much wider variety of applications. [ABSTRACT FROM PUBLISHER]