Exposing Engineering Data As A Strategic Asset

Exposing Engineering Data As A Strategic Asset

The Air Force SEEK EAGLE Office (AFSEO) has developed a new data lake architecture and applications that combine historical data, machine learning, and modern big data technology to accelerate complicated engineering analyses. Before a new store (weapon, fuel tank, etc.) is mounted on an aircraft, AFSEO is responsible for setting safe limits for flight for that store and determining the unsafe ways in which different stores can interact. Doing so requires complicated engineering analyses and tests (flight, wind tunnel, etc.) led by PhD level engineers. As AFSEO surpasses 50+ years of accumulated data, the promise of finding, using, and synthesizing previously approved analyses rather than building a new analysis from scratch is too large to ignore. The data lake addresses this problem in two ways. First, a custom data catalog has been built that tags all files with relevant metadata and then makes them searchable and filterable through a web-based UI. This is enabled by a Dell EMC multiprotocol Isilon storage solution that simultaneously serves files to all operating systems based on users’ roles through SMB/NFS protocols and to the Cloudera based data lake as HDFS. The custom data catalog UI is hosted and served directly by Solr, offering a number of advantages. Second, a custom application built on Tamr machine learning searches historical documents for relevant antecedents, and in some cases, can create a “by analogy” certification without human intervention. Both efforts help engineers certify new stores and capabilities with the agility required to support the modern-day US Air Force

Event Details
  • Start Date
    September 22, 2020 2:00 pm
  • End Date
    September 22, 2020 3:00 pm
  • Status
    ARCHIVED
  • Location
  • Organizer