Improving life cycle management through simulation and efficient design

Abstract

Life Cycle Management (LCM) is defined as a decision-making process that takes into consideration the benefits, costs, and risks associated with each action over the full life cycle of a system. Effective LCM requires good forecasting to help determine future requirements for design and development, acquisition, in-service support and sustainment, modernization, and final disposal of a fleet of systems. It is in forecasting that simulation tools play a key role in LCM by helping program managers to gain insights into their supported systems. The Total Life Cycle Management Assessment Tool (TLCM-AT) is a probabilistic modeling and simulation analysis tool developed to support and improve the USMC's LCM. This powerful tool is capable of performing "what-if" scenario analysis to compare the merits of multiple courses of action (COAs) or policies. Unfortunately, such analytical results are predicated on a set of conditions developed in the model that have little chance of occurring in real life. This thesis introduces a Java-based application that combines the capabilities of TLCM-AT with the benefits of a sophisticated design of experiments (DOE) to perform in-depth sensitivity analysis of alternatives. A well-developed DOE can simulate real life by modeling a wide range of conditions under which the performance of each COA is measured. Data from this kind of experiment can be used to help in the development and selection of robust COAs and policies.