Analytic framework for TRL-based cost and schedule models

Abstract

Many government agencies have adopted the Technology Readiness Level (TRL) scale to help improve technology development management under ever increasing cost, schedule, and complexity constraints. Many TRL-based cost and schedule models have been developed to monitor technology maturation, mitigate program risk, characterize TRL transition times, or model schedule and cost risk for individual technologies as well technology systems and portfolios. In this thesis, we develop a 4-level classification of TRL models based on the often-implicit assumptions they make. For each level, we clarify the assumption, we list all supporting theoretical and empirical evidence, and then we use the same assumption to propose alternative or improved models whenever possible. Our results include a justification of the GAO's recommendations on TRL, two new methodologies for robust estimation of transition variable medians and for forecasting TRL transition variables using historical data, and a set of recommendations for TRL-based regression models.