Reliability improvement of major defense acquisition program cost estimates—Mapping DoDAF to COSYSMO

Major defense acquisition programs (MDAPs) are notoriously prone to substantial cost overruns, and engineering and design issues are often partially to blame. Accordingly, as the Department of Defense (DoD) develops increasingly complex systems, additional emphasis is being placed on systems engineering to ensure that programs remain on schedule and within budget. Notably, in 2013 the DoD mandated the submission of DoD Architecture Framework (DoDAF) models during a program’s Materiel Solution Analysis Phase—early in the system life cycle—a period historically characterized by a lack of system specification and parametric estimation. Capitalizing on this change, this paper establishes an explicit connection between the DoDAF and the Constructive Systems Engineering Cost Model (COSYSMO) with the goal of improving the measurement reliability of MDAP cost estimates. In particular, we employ techniques from text mining and social network analysis to produce a mapping between DoDAF’s 52 models and COSYSMO’s 18 drivers, yielding an untapped source of information for determining functional system size and, ultimately, estimating systems engineering effort.