An Efficient Approximation for Project Composite Learning Curves

The project manager is responsible for providing an accurate project schedule estimate to all parties involved in a new project. To improve the accuracy of such schedules, especially for programs consisting of repetitive projects, it is necessary for the program/project manager to incorporate the “learning effect” into activity time estimates. Any activity that a worker will be performing repetitively will likely see its duration time decrease along a predetermined learning curve. To ignore the learning effect during the estimating phase is to settle for inaccurate estimates and likely broken delivery promises. While previous authors have argued the wisdom of incorporating learning curves, they have only provided computationally expensive means of doing so. This paper demonstrates through several diverse examples of multiunit programs that project composite learning curves can be approximated with minimal effort to provide accurate estimates of program duration and delivery dates.