Anticipating High Demand Class IX Supplies in the Western Pacific Area of Operations

Abstract

The thesis focuses on predicting high-demand class IX items for MARFORPAC by analyzing maintenance and requisition data over multiple fiscal years. We apply a Markov chain model to examine failure patterns across subassembly systems of a specific assets; this model will provide a descriptive analysis to examine the maintenance history by identifying the primary defects that have caused failures within a given major end item. We will then apply this model to help us understand how the probabilities of one failure in a subassembly can lead to failures in another, allowing us to predict future breakdowns based on historical data. By identifying these failure sequences, we aim to provide a framework for MARFORPAC units to leverage predictive analytics in distributed and contested environments.