Date of Award

3-23-2017

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Systems Engineering and Management

First Advisor

John Elshaw, PhD.

Abstract

Learning curves are used to describe and estimate the cost performance of a serial production process. There are numerous different models and methods, however, it is not definitively known which is preferred. The research objective is to performance compare the more common learning curve models. The research goal is improved understanding of the systemic cost drivers of a production process, their relationship to cost, and present modeling methods. The research method is qualitative analysis combined with statistical regression modeling. The research identified that preference for one function or another depended upon the shape of the data and how well a model formulation could be made to fit that shape. This depended upon the models basic shape and the available parameters to alter its appearance. The typical learning curve model assumes that cost is a function of time but commonly omits factors such as production process resources changes (capital and labor) and its effect on cost. A learning curve model that includes the effects of resource changes would likely provide higher estimative utility given that it establishes a systemic relationship to the underlying production process. Additional research and data is required to further develop this understanding.

AFIT Designator

AFIT-ENV-MS-17-M-193

DTIC Accession Number

AD1055207

Share

COinS