Posted by

Estimating Cost To-Go Without Stable EVM Data

Earned Value Management Track




Initial independent estimates for development programs are typically created using parametric methods. As a program progresses, it becomes important to separate sunk cost from cost to-go. In the very early stages of a program, it is acceptable to estimate cost to-go by simply subtracting the reported sunk cost from the parametrically-estimated total cost. However, as the program continues to progress and more significant progress accumulates, we can no longer assume that dollars spent equate directly to technical progress. It is then necessary to understand progress to-date in order to estimate cost to-go. If a detailed, stable, trusted, schedule is available, one could potentially estimate costs to go by independently estimating the cost of each of the incomplete activities remaining in the schedule. However, parametric CERs at the detailed schedule activity level are seldom available and independent estimating teams seldom have access to the detailed information required to do build-up estimating at that level. A good solution to this problem is to use earned-value data to make an independent assessment of technical progress, and then apply this information to the parametrically-estimated total cost to determine cost to-go. This works well if the earned-earned value data is stable and trusted to represent a stable baseline. If the earned-value data is not stable, a new approach is required. This paper examines a sample program in which the earned value data was obviously not stable, despite the Critical Design Review having been accomplished and a significant amount of money having been spent. In this case, the earned value data indicated that on the average over several years, management’s perception of the total scope of the development effort grew as fast as work was accomplished. Although this growth was described by management as “requirements growth,” the technical performance parameters to which the parametric CERs responded did not indicate growth in performance requirements. It was apparent that it was not the performance requirements that had grown but management’s understanding of the difficulty of the effort required to meet the requirements. As such, the earned-value data could not be trusted as an indicator of relative technical progress. A new approach was required.

This paper describes an approach to determining progress to-date and cost to-go based on major milestone completion dates. Specifically, we looked at the percent of total actual dollars spent before the CDR milestones for subsystems in each of several analogous programs. We then used the amount of time before and after CDR for each subsystem in the system of interest to determine completion percentages for each subsystem. This percentage, applied to the parametrically-estimated total costs, provided a credible independent estimate of cost to-go.



Peter Frederic
Tecolote Research, Inc.
Mr. Frederic has been with Tecolote Research since 1983. He is the Chief Scientist of Tecolote’s Santa Barbara Group, which includes offices in Santa Barbara, Ogden, Albuquerque, and Dayton. In his career at Tecolote, Mr. Frederic has logged a wide variety of experience including cost database software development, CER development, technical baseline development, schedule analysis, simulation software development, and cost estimating. His efforts have addressed many aerospace technologies including radars, optical sensors, missile systems, launch vehicles, launch facilities, space vehicles, and aircraft.

Ronald K. Larson
Mr. Larson joined NASA in 2003 and has served in both the Cost Analysis Division and the Independent Program Assessment Office (both elements of the NASA Office of Evaluation) where he is a Technical Resource Analyst responsible for performing independent programmatic assessments on major NASA projects as they reach key decision points and major milestones. Prior experience was gained working for the Aerospace Corporation, The MITRE Corporation, MCR, Inc., Astrium GmbH, Lockheed Missiles and Space Company, The Chrysler Corporation Space Division and the United States Air Force where he began his career as an Atlas F launch officer.