Interactive model trading for resilient systems decisions

Abstract

The Interactive Model-Centric Systems Engineering research effort is interested in developing knowledge necessary to leverage the increasing involvement of computational models in system design. A key activity in such modelcentric environments is the selection and usage of models to generate data for decision making. Extending work from prior demonstrations, this paper presents a case study where insights are gained via usage and comparison of multiple models. The results highlight the need to explore how model choice and trade-off can impact the attractiveness of alternative systems. Sixteen tradespaces of cost-benefit data are generated via combined pairwise usage of four alternative evaluative models (performance and cost calculations) and four alternative value models (calculating the "goodness" of different levels of performance and cost). These tradespaces are used to determine attractive Pareto efficient Space Tug vehicles, as well as insights that are less sensitive to model choice. No best model is shown, but rather different models provide insights into different aspects of the system evaluation and valuation activity. Two categories of insights are highlighted: patterns in the structure of the decision problem (i.e., how "value" is defined and what systems might be feasible) and artifacts of the models themselves (i.e., how to mitigate against misleading results due to model abstractions).