Improving System Sustainment through an Integrated Modeling Schema Coupled with Effective Execution of the Lifecycle Sustainment Plan

Abstract

"Title 10 § 4324 tasks the Product Support Manager (PSM) to “(B) ensure the life cycle sustainment plan is informed by appropriate predictive analysis and modeling tools that can improve material availability and reliability, increase operational availability rates, and reduce operation and sustainment costs.” Advances in modeling and simulation offer the opportunity for PSMs to bring new approaches to long-standing challenges, particularly using AI and machine learning models. This paper examines how one PSM has used a series of traditional and AI-based models to develop predictive analytics that can advise the platform life cycle with the expectation of improved Operational Availability (Ao) and Material Availability (Am). There are many challenges, including: (1) most models are built to suit the particular user community, without any intention of connecting the model to others, (2) each model is often built with a set of algorithms that are custom adapted to the problem set, giving rise to composability questions, and (3) many models are built to different time scales, or even independent of any time representation. Life cycle sustainment of submarines, particularly during service life extension, has been met with challenges that have led to inefficient use of time and personnel resources. While maintenance availabilities include various service, planned, corrective and alteration jobs that maintain or increase readiness of the Navy’s deterrent fleet, these facilities encounter cost and schedule overruns caused by constraining factors including personnel, equipment, facilities, supplies, material, weather, or other uncontrollable factors. The COLUMBIA Submarine Program has developed several models to assist in decision making. We describe two models, one a discrete event simulation of the approved and alternate life cycles and the other a manpower forecasting model for the repair facilities and how these models have led to new insights in improvements that will improve Ao and Am. We also describe a future state where currently disconnected models are integrated together, allowing decision makers insights to see the complete loop from a 3D product model used to design, build, and sustain the platform to the end user applications."