Framework for Military Aircraft Fleet Retirement Decisions

Abstract

The purpose of this work is as follows. Military aircraft are enormous investments for a nation. The systems lifecycle for aircraft spans decades wherein aging effects increase maintenance and operations costs over time. At some point, the deterioration of a fleet of aircraft erodes the capability of those assets below an acceptable threshold, thus triggering retirement planning by a military. Questions arise about how to retire a fleet, including how many aircraft should be retired, when those aircraft should be retired and which aircraft should be chosen. There are few military aircraft fleets that are retired each year, and even fewer managers who understand the aircraft retirement puzzle. This work addresses these questions. The purpose was to provide fleet managers with a comprehensive framework to guide decision-making, as well as to build tools and a standard guidance framework for fleet managers to implement.

In terms of methodology, in the absence of directly applicable existing research in this field, fleet management concepts and modelling approaches were studied in related fields and then applied to the military fleet retirement problem. The vital first approach to the problem required the baselining of military aircraft fleets given structural loading data and utilization histories. Database analysis and trending algorithms were written to draw correlations between existing data and structural fatigue effects. This work then implemented a greedy algorithm model to solve the individual aircraft retirement scheme. That led to a mixed-integer linear programming approach to optimize a fleet utilization and rotation model. Combined, these methods provided concrete steps for the fleet retirement decision framework, which followed established methods for designing a decision support framework. Throughout the work, a consistent case study fleet (United States Air Force’s A 10 Thunderbolt II) was utilized to provide validation of the methods, while secondary case studies and validation techniques were employed to test applicability of the methods to other military aircraft fleets and other capital asset types.

In terms of concrete research results from the work carried out, this dissertation discovered that a framework for military aircraft fleet retirement decisions was a needed contribution to the field. In the process of building that framework, other valuable results were obtained. It was found that aircraft utilization information could be correlated to cyclic loading data on an individual aircraft level. This revealed patterns in aircraft fleets showing which mission types and basing locations either increased or decreased structural degradation. Using that information led to the result that a fleet manager could determine which aircraft to retire prior to others while optimizing an objective function related to fleet cost, fleet utility or the ratio thereof. It was also found that a fleet manager could selectively utilize individual aircraft at particular bases flying particular missions to prolong or hasten the structural degradation of those aircraft. This led to the result that a fleet manager could therefore forecast retirement dates for an entire fleet, subpopulations within that fleet or individual assets.

From the research carried out, it is emphatically concluded that the results imply that a fleet manager beginning with only aircraft usage data can actively manage a fleet of aircraft to extract residual value from the fleet prior to retirement. This work showed that resource allocation could be improved by utilizing a mixed integer linear program to schedule asset retirements. Further, this work illustrated how a management strategy could impact future usage levels in a way to extend useful lifetime. With a capital asset as critical to national defense and as expensive to acquire, operate and retire as military aircraft, focusing on the end-of-life phase of the systems lifecycle not only promotes forward thinking but also provides potential cost savings. This work’s limitations included its focus on military aircraft instead of all capital assets and that the methods were not implemented in an actual fleet environment. This dissertation demonstrated that a flexible framework with core modelling elements is a tool capable of solving the problem of aircraft fleet retirement decisions. Fleet managers both military and otherwise should investigate the applicability of the methods and findings in this dissertation to their own challenges. Future research must include application of the methods to an actual operating fleet. Also, the methods should be applied to other capital asset classes including military equipment and commercial equipment.