Date of Award

3-24-2016

Document Type

Thesis

Degree Name

Master of Science in Engineering Management

Department

Department of Systems Engineering and Management

First Advisor

Vhance V. Valencia, PhD.

Abstract

Mission Dependency Index (MDI) is a metric developed to capture the relative criticality of infrastructure assets with respect to organizational missions. The USAF adapted the MDI metric from the United States Navy’s MDI methodology. Unlike the Navy’s MDI data collection process, the USAF adaptation of the MDI metric employs generic facility category codes (CATCODEs) to assign MDI values. This practice introduces uncertainty into the MDI assignment process with respect to specific missions and specific infrastructure assets. The uncertainty associated with USAF MDI values necessitated the MDI adjudication process. The MDI adjudication process provides a mechanism for installation civil engineer personnel to lobby for accurate MDI values for specific infrastructure assets. The MDI adjudication process requires manual identification of MDI discrepancies, documentation, and extensive coordination between organizations. Given the existing uncertainty with USAF MDI values and the effort required for the MDI adjudication process, this research pursues machine learning and the knowledge discovery in databases (KDD) process to identify and understand relationships between real property data and mission critical infrastructure. Furthermore, a decision support tool is developed for the MDI adjudication process. Specifically, supervised learning techniques are employed to develop a classifier that can identify potential MDI discrepancies. This automation effort serves to minimize the manual MDI review process by identifying a subset of facilities for potential adjudication.

AFIT Designator

AFIT-ENV-MS-16-M-184

DTIC Accession Number

AD1054119

Share

COinS