Empirical Cost Estimation Tool (PMS 320)

Abstract

This research project pertains to the identification, review, and potential development of existing and alternative ship cost modeling methodologies. Most ship cost modeling has been traditionally weight-based. This approach drives the U.S. Navy decision makers to acquire smaller ships that require custom-designed shipboard components. Current, and future, Department of Defense (DoD) acquisition budgeting processes require identifying, modeling, and estimating the costs of shipbuilding. The purpose of this research project is to determine if there is a more accurate way to empirically predict and model ship acquisition costs. The cost modeling tool developed in this study is intended to support development of ship cost forecasts. The proof of concept example for using this cost modeling tool included herein will provide a roadmap for other new ship acquisition cost modeling. The outcome of this research will likely increase cost savings. The focus of this research is a comprehensive review of the most promising cost modeling methodologies. Notional cost data, or rough order of magnitude values, will be collected or generated to support this review of the cost methodologies. These data will be generated by the researchers using archival cost data from ship maintenance projects of various destroyer (i.e., DDG) acquisitions. We will identify these extrapolations, and we will use the resulting notional data to help evaluate the efficacy of the various cost models. This approach allows readers and study sponsors to see the various types of cost models, approaches, and sample data variables that are required to run the cost models and to examine sample results, as well as review the pros and cons of each approach. This study may require a follow-on project if there is a method that is of interest or that the sponsors feel might be applicable for a given ship acquisition context. The required data variables as well as sample results will be listed in the report, so the sponsors will know what to expect prior to engaging in any new research project. A follow-on study would allow us to obtain real-life cost data that could be plugged into the desired cost model. The selected cost model will likely include the standard parametric models, nonparametric methods, systems dynamics based on project management task-based schedule and cost models; semiparametric Monte Carlo simulation models; curve fitting, time-series, and cross sectional models; nonlinear models, and so forth that have proven useful in forecasting costs in other acquisition contexts.