Learning curve application to space shuttle processing simulations

Traditional learning curves were pioneered by T.P. Wright in 1936, with the idea that improvements in labor-hours to manufacture an airplane could be described in a mathematical pattern. This paper will show that this concept of learning curve improvements to production metrics can be applied based on cumulative time, rather than volume of production, for one-of-a-kind applications, such as space shuttle flights, where production quantities are very limited. Policy and process changes can also be observed in production data, and the learning curve is useful in the prediction of future trends. Past data from space shuttle processing is demonstrated to fit this new definition, and prediction of future process metrics is explored. Once the learning curve is time-based, simulation can be applied to model the system and enhance the prediction effort for future process metrics.