An Experiment to Improve Cost Estimation and Project Tracking for Software and Systems Integration Projects

It is becoming increasingly difficult to predict the resources, costs, and time scales for the integration of software and systems built from a number of components supplied by third parties. Many cost models use the concept of product size as the prime driver for cost estimation but our experience has shown that the supplied quality of components and the required quality of the integrated systems are becoming the dominant factors affecting the costs and time scales of many projects today. ICL has undertaken an experiment using an alternative life cycle model, known as the Cellular Manufacturing Process Model (CMPM), which describes how a product is integrated from its constituent components. The experiment has helped to develop a method and a set of metrics to improve cost estimation and project tracking.