Dimensioning A Product Development Project Portfolio Using A Closed Queueing Network

New product development processes are difficult to model as they are subject to high variability and are hard to monitor. In multi-project environments, one particular problem is to quantify the effects of changing the number of projects that run concurrently. In this paper, a closed queueing network model of the flow of projects in a product development system is presented. The model is evaluated using discrete-event simulation. Using a relation between a project's monetary value and the time it spends in the system, an optimal number of concurrent projects is determined. Application of the model and results are shown for a case study of a Danish biotechnology company. Results show, that the value generated by the system can be increased by 17.5 percent by adjusting the total number of concurrent projects and by terminating low-value projects midway. Though the model is tailored to a specific system, it contains many features relevant for new product development systems in general.