Scenario Set as a Representative of Possible Futures

Abstract

Because of the inherent disconnect between the length of the procurement cycle and the dynamic time scales of the security environment, future military capability planning presents a unique challenge. The security environment can change very rapidly, and the extent of the changes, and thus the future environment, is potentially unpredictable. In contrast, the procurement cycles responsible for the renewal of the capabilities are typically on the order of decades. The traditional approach to capability-based acquisition consists in the development of a scenario set representing possible futures, and then plan capabilities to address the challenges presented by the scenario set. This approach is based on an implicit assumption that the chosen scenario set is sufficiently representative of all possible futures. Intuitively, if the scenarios are selected somewhat randomly (i.e., there is no significant selection bias that would eliminate certain futures), then a solution that works for the set of planning scenarios, whether in optimization or capability scheduling sense, is more likely to work for any feasible future. In this paper we examine this assumption in a more systematic way and attempt to express likelihood that a solution to a limited set of vignettes is a solution for all possible futures, at least for some special cases.