Uncovering the fragility of large-scale engineering projects

Abstract Engineering projects are notoriously hard to complete on-time, with project delays often theorised propagate across interdependent activities. Here, we use a novel dataset consisting of activity networks from 14 diverse, large-scale engineering uncover network properties that impact timely completion. We provide empirical evidence perturbation cascades, where perturbations in the delivery single can up 4 activities downstream, leading large cascades. further show clustering significantly affects overall delays. Finally, find poorly performing have their highest high reach nodes, which lead largest while well low resulting localised Altogether, these findings pave way for network-science framework materially enhance projects.