Causal Factors of Near Misses and Accidents in Urban Railway Construction: A Complex Network Approach

Near misses occur much more frequently than accidents, often by factors of hundreds or thousands, and they typically share common causes with accidents. It is believed that eliminating the causes of near misses can make a significant contribution toward accident prevention. However, existing studies on causal factor analysis have mainly focused on accidents, with little attention paid to near misses. To address this gap, this study aims to explore causal factors of near misses and accidents in urban railway construction using a complex network approach. A two-layer urban railway construction safety risk network (URCSRN) was established based on the chain relationships in near misses and accidents, and its topological features were analyzed and revealed. Besides, the network robustness assessment was conducted through random and deliberate attacks. Results showed that both the accident layer network and the near-miss layer network in the URCSRN were scale-free networks. A positive degree correlation existed between the accident layer network and the near-miss layer network. A total of 11 critical factors leading to near misses and accidents were identified and found to exhibit spatial-temporal and worker typer-related characteristics. This study is beneficial for revealing the same contributing factors of near misses and accidents and the occurrence patterns of these factors among diverse worker types, in different construction phases and areas. It can provide practice guidance for safety managers to take safety measures to prevent similar near misses and potential accidents. Also, this study offers empirical evidence to support the common cause hypothesis and improves accident causation theory. Using a multilayer network to examine near misses and accidents simultaneously provides a new perspective for construction incident analysis.