Article ID Journal Published Year Pages File Type
488761 Procedia Computer Science 2015 10 Pages PDF
Abstract

With the rise of readily accessible computation and data storage capabilities, models have been increasingly leveraged for generating system data to support decision making. Moving beyond considerations of what makes a good model, Interactive Model-Centric Systems Engineering seeks to address core challenges as well as key implications associated with relying upon models for generating the data upon which important system decisions are made. Among these is the issue of understanding the consequences of model choice itself. This paper proposes three categories of models: performance, cost, and value, and addresses the question of value model trading. The concept is illustrated through a demonstration Space Tug satellite system case where four value models (multi-attribute utility, analytic hierarchy process, cost-benefit analysis, and measure of effectiveness) are each used to assess the goodness of potential system alternatives. The consequence of value model choice changes the Pareto efficient set of “best” system solutions, and a comparison of these sets reveals the potential of finding system solutions that are robust to value model choice, supporting the strategic goal of identifying resilient system decisions.

Related Topics
Physical Sciences and Engineering Computer Science Computer Science (General)