Application of Data Mining Techniques to Air Vehicle Systems and Software

Description

Lockheed Martin has a large number of flight discrepancy reports and software problem/anomaly reports describing issues reported in software design and development and in aircraft operations.  However the volume of this data can overwhelm existing analysis techniques performed primarily manually.  Initial results from our project included a feasibility assessment.  DePaul researchers have developed a customized algorithm for detecting cross-cutting and reoccurring problems in the flight discrepancy reports.  Results have been presented to Lockheed Martin affiliates.  The project is now moving to the next phase which involves developing a customized tool to enable engineers to review results and to provide feedback.