Using Pre-Milestone B Data to Predict Schedule Duration for Defense Acquisition Programs

Abstract

Accurately predicting a realistic schedule for a defense acquisition program is a difficult challenge considering the inherent risk and uncertainties present in the early stages of a program. Through the application of multiple regression modeling, we provide the program manager with a statistical model that predicts schedule duration from official program initiation, which occurs at Milestone B, to the initial operational capability of the program’s deliverable system. Our model explains 42.9% of the variation in schedule duration across historical data from a sample of 56 defense programs from all military services. Statistically significant predictor variables include whether a program is a new effort or modification to an existing program, the year of Milestone B start as it relates to changes in defense acquisition reform policy, and the amount of raw funding (adjusted for inflation) prior to Milestone B for a program. Our final and strongest predictor variable, percentage of the total RDT&E (Research Development Test and Evaluation) funding profile allocated at Milestone B, indicates that increased percentage of RDT&E funding for pre-Milestone B technology risk reduction may shorten a program’s schedule duration to initial operational capability.