Capturing risk in solution prioritization.(contemporary issues)(Report)

Introduction

The intent of the Chief of Staff of the Air Force-directed Comprehensive Assessment of Nuclear Sustainment (CANS) was to identify and provide solutions for any systemic problems within the nuclear sustainment enterprise. During the study, strategic-level findings were prioritized using a multi-objective optimization approach. As with any system, there is risk associated with maintaining the status quo when problems are identified. However, as it was assumed that the risk associated with not addressing the strategic-level findings was sufficiently high such that all findings would eventually be addressed, this risk was not included in the prioritization problem formulation. This is not to say that risk was not considered at all, but rather that the consideration of risk was limited to that of the solutions themselves--risk that the solution may have unintended consequences that actually make the problem worse. Though sufficient and appropriate for the initial study, the inclusion of risk in the prioritization formulation may provide a more complete picture for decisionmakers. The remainder of this article presents a generic methodology to incorporate a measure of risk into a multi-objective solution prioritization problem like the one in the CANS study. Such a methodology may be useful in follow-on efforts of CANS. (For a detailed description of the CANS methodology, see the previous Air Force Journal of Logistics article entitled "Using AFSO21: The Problem is Big, Time is Short, and Visibility is Enormous." (1))

Original Problem Formulation

The original prioritization portion of CANS attempted to prioritize strategic level findings of the study. To accomplish this, subject matter experts (SME) scored the impact of each strategic level finding, if solved, on the five key mission areas (see Figure 1). The result was then formulated as the following multi-objective optimization problem:

max F(x)

subject to

[chi][member of][OMEGA] = [{[chi] [member of]{0,1}.sup.n] : [g.sub.i]([chi]) [less than or equal to] 0, i = 1,2, . , M}