Product Obsolescence Prediction (POP) Tool

Predicting and managing obsolescence is a challenging task that requires balancing operational needs with supply chain realities. To enhance the DLA’s ability to manage obsolescence, our team will define, design, and implement a Product Obsolescence Prediction (POP) tool powered by advanced AI and ML methodologies. These AI/ML technologies will provide predictive capabilities, analyze the complex web of sourcing interdependencies, and automate currently manual and often reactive obsolescence management processes. Innovative AI/ML methodologies will be leveraged to scrutinize a comprehensive array of data sources such as from design engineering, field/ops, maintenance, repair & overhaul, industrial/trends, geopolitical, and economic/financial, to identify parts, software, or equipment at risk of becoming obsolete; and the tool will be scalable to include new data as future sources become available.  POP's AI/ML technologies include using generative AI (GenAI) to extract information from multiple sources of data to enrich predictions of part obsolescence.  Our POP tool will also contain a machine learning operations pipeline that has a hierarchical system of machine learning models, where models earlier in the pipeline help enrich and inform the final model for part obsolescence scoring. AI/ML techniques will be used to simulate different scenarios of part obsolescence and assess their impact on DLA's operational capabilities and readiness to assist in mitigating issues for part availability for the different scenarios. This proactive approach empowers the DLA to make informed supply chain decisions that improve sustainment, modernization, and lifecycle management, optimizing resource allocation and ensuring mission readiness. Our POP tool will also integrate ML algorithms with knowledge graph systems, enhancing the DLA’s ability to assess supply chain vulnerabilities, predict obsolescence, and visualize sourcing patterns.  Knowledge graph systems improve entity resolution, and the POP tool will be able to assist with configuration control across different systems and platforms.  While not all challenges can be predicted, the synergy between AI/ML technologies and human expertise improves decision-making processes, allowing DLA to proactively address obsolescence and mitigate budgeting challenges. Adopting POP's AI/ML methodologies will give the DLA a significant advantage in predicting obsolescence and better managing the DoD’s extremely complex logistical environment.