R. Bansal, P. Ren Toh, T. Raheja, R. Ashok, V. Hooda, S. Panhalkar, A. Tambe
Kipo AI, California, United States
Keywords: sustainment, equipment, obsolescence, supply chain, electronics
Critical assets, including defense equipment and public infrastructure, must be actively maintained over long, 30-year lifecyles. Because critical components in these assets may have lifecycles shorter than the asset itself, they need to be replaced to maintain the asset. Unfortunately, replacement components may become unavailable from their original manufacturers due to supply chain disruptions or obsolescence. Traditional methods of selecting replacement components require trained engineers to manually inspect "datasheets" outlining the technical specifications of replacement components. This is a highly time-consuming process that takes several weeks to find appropriate substitutes and is prone to human error, rendering critical public infrastructure and defense equipment unusable during this time. We exhibit a software platform, Kipo AI, that tackles this problem. Kipo AI aggregates technical and supply chain data for millions of off-the-shelf components and uses Large Lanugage Models to standardize this data. Leveraging search algorithms over the standardized data that match the form, fit, and functionality of two components, Kipo AI entirely automates the process of selecting replacement components, accelerating maintenance cycles by an order of magnitude. Kipo AI enables organizations to rapidly respond to obsolescence and supply chain disruptions that would otherwise prohibit them from maintaining long-lifetime assets.