Enhancing dynamic resilience of traffic network: Role of Bayesian pre-posterior analysis supported by real-time congestion scanning technology

• Dynamic resilience capability of disrupted traffic network is advanced. • Dynamic stochastic assignment method-based traffic control system is applied. • Bayesian pre-posterior method supported by traffic scans used as a decision aid tool. • A bridge outage case study is used to test the methodological framework. • Suggests dynamic resilience action tested in digital twin for emergency preparedness. Major unanticipated disruptions to the traffic network that result in severe delays to travellers and adverse socio-economic impacts require an emergency preparedness strategy. Examples are bridge collapse caused by collision with a ship or fire or structural failure, and bridge closure due to stormwater. Although the inherent resilience of the traffic network can reduce the intensity of effects, dynamic resilience is required to further reduce delays until the disruption is removed or necessary demand management actions are implemented. The emergency preparedness mandate can be well-served with dynamic resilience action tested in a digital twin of the traffic network. This paper reports research on the development of dynamic resilience capability of the traffic network. Following problem definition, the methodological framework is described that incorporates dynamic stochastic assignment method-based traffic control system, real-time congestion scanning technology, and Bayesian pre-posterior analysis method that requires the use of scanning technology. The developed methodological framework guides decisions on the choice of dynamic resilience action and implementation time to minimize delay. The Chaudière Bridge that links Cities of Ottawa and Hull/Gatineau in the Canadian National Capital Area is used as a case study. This bridge outage caused by stormwater demonstrates the role of dynamic resilience in addressing uncertain states of post-disruption traffic delay. As compared to the business-as-usual traffic control, the dynamic resilience action-based control reduces delay by 12.3% under very high delay condition. The developed new methodological framework for analyzing network resilience can be applied to other major network disruption cases to reduce delay.