Ship Construction Estimates at Completion: A New Technique Using the Weibull Function
EVM & Scheduling Track
This presentation summarizes recent earned value management (EVM) research at Technomics, Inc. exploring new, improved methods for generating Estimates at Completion (EAC). Our approach is based on a significant amount of “burn rate” data resulting from serial ship procurement. We define burn rate as the incremental change in the cumulative value of Actual Cost of Work Performed (ACWPCum) divided by the time over which the change occurred. Though some research suggests that Rayleigh modeling of the burn rate is preferred, we have studied an alternative regression technique based on the Weibull probability density function, known as the Hybrid Weibull EAC Model. The Hybrid Weibull EAC Model is a piecewise equation consisting of four parts: three separate Weibull curves and a linear adder based on average burn rate.
A common problem associated with nonlinear regression is the lack of data. When limited data are available, regression equations with multiple parameters are susceptible to generating unrealistic parameters that could severely underestimate or overestimate the ACWPCum at program’s end. To counter this problem, we use the data from early ships (“own class” data) to determine and to apply fact‐based constraints for later ships such that the parameters generated by the regression equation are realistic and yield more accurate Estimates at Completion. The Hybrid Weibull EAC Model uses two constraints: the time and amplitude of the peak burn rate. Since the profile of burn rates across all investigated production items is similar, the time and amplitude of peak burn rates for items in the early stages of procurement can be extrapolated.
This presentation provides overview of how this methodology is employed.
Shawn Rudolph is a Cost Analyst for Technomics, Inc. He graduated from Virginia Polytechnic Institute and State University in May of 2008 with a B.S. in Mathematics and a minor in Statistics. His coursework included Theoretical Statistics, Applied Time Series, and Advanced Calculus. After graduation, he worked as an Economist for Tesla, Inc. where he recalibrated econometric models that forecast electric load for various utility companies. Since joining Technomics, Inc. in early 2009, he has conducted cost research for the Air Force Cost Analysis Agency (AFCAA) and has provided Independent Estimates at Completion for the VIRGINIA Class Submarines in support of NAVSEA 05C.