A Distributed Bayesian Network Approach for Risk Modeling and System Sustainment.
Abstract
In this proposal, the ISE department will work with NUWC Keyport; specifically, system developers in Code 465. The group develops and supports the Obsolescence Management Information System (OMIS™) which in turn supports the work of program managers. To give decision-makers more lead-time to develop plans to support acquisition and sustainment, the proposed work will develop a distributed set of network models and integrate those models to advance supply chain risk management for long-life systems in the DoD. The proposed work builds upon prior research that developed predictive analytic algorithms for predicting lifecycle phases at the part level specifically for COTS, electronic and mechanical parts. The results have been published in refereed journals and presented at conferences. The code for all algorithms has been transitioned to developers. While the prior work focused on utilizing predictive analytics for risk at the part level, the proposed work takes a broader, system sustainment view of the problem and will encompass supply chain risk management (SCRM). SCRM embodies several risk systems: macro risk, demand risk, manufacturing risk, supply risk, infrastructural risk. As such the proposed research has two foci: • Development of distributed Bayesian network models for the appropriate risk system (i.e. from the list above). • Distributed Bayesian network reconstruction for SCRM monitoring and sustainment state prediction of the system.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 12, 2025
- Source ID
- N001742310013
Entities
People
- Christina M. Mastrangelo
Organizations
- United States Navy
- University of Washington