A multiobjective optimization method considering process risk correlation for project risk response planning

Abstract

Project risk response planning is an important component of project risk management to help ensure project success. However, thus far, the relevant research has ignored the correlations between risk factors during different project development subprocesses. This study proposes a novel risk response method by opening the black box of the project process and considering the risk correlations among different subprocesses to help practitioners create a practical risk response plan. In our method, we first identify process risk factors from two-dimensional criteria called risk categories and project development subprocesses. Then, we consider the risk correlations between different subprocesses to assess the true effect of each process risk factor. Finally, we build a multiobjective risk response model under resource constraints to minimize the total expected losses, the total expected schedule delay, and the total expected quality reduction. Based on the model, we obtain a set of Pareto-optimal risk response plans. Project managers can select a satisfactory risk response plan based on their experience and objective project requirements. Finally, we apply the proposed method to a real software development project. We find that the risk response effects are better if managers control the process risks earlier. Moreover, the number of initial risk factors in each subprocess is an important reference in risk response resource allocation.