Modeling and analysis of cascading failures in projects: A complex network approach

Abstract

Abstract Due to the complex integration of the tasks in a project, the failure of a single task can trigger extremely large-scale failures and destroy a considerable part of the overall project. To investigate cascading failures in projects, in this paper, the project is first abstracted as a weighted directed network composed of tasks and task interactions, after which a cascade model that takes account of the project’s self-protection mechanism is developed to examine a failure propagation process originates from a single task failure. The model is then applied to examine cascading failures in a project under three types of single task failures with varying parameters. The experiment results demonstrate that the method is very effective in predicting cascading failures, evaluating the impact of cascading failures on the project, and identifying large cascades. The insights gained from the simulation results have implications for managers in the implementation of protective measures to mitigate cascading failure risk and avert project catastrophes.