In 2020, the National Infrastructure Commission  announced the regulatory obligations for developing resilient infrastructure systems in UK, where it defines “resilience” as the capacity of infrastructure systems (IS) to anticipate, absorb, resist, recover, adapt and transform any disruptive forces (e.g., natural disasters, extreme weather) [2, 3]. Along these rising concerns, a continued increase in the frequency and intensity of the climate-induced natural disasters (e.g., flash floods and extreme weather) have made them the most critical risks to the IS . Moreover, in addition to the disruption caused by natural disasters to an individual system’s serviceability, the inter-connectivity of IS causes several challenges for their recovery [5, 6]. As an instance, disruptions in transportation systems reduces the accessibility of other IS and further impedes their recovery. Accordingly, there is an immediate need (1) to enhance the capacity of existing IS for absorbing and resisting climate-induced disasters; and (2) to develop an optimal recovery plan for retrieving the service conditions of IS in post-disaster conditions, in which the inter-connectivity of the systems is properly addressed. This research project aims to address the latter need by developing a data-informed decision support system to optimize the recovery process of IS in post-disaster conditions.
The resilience of IS is largely affected by the risk management practices applied during their lifecycle [2, 7], though managing risks in these projects faces several challenges due to: (1) numerous uncertainties associated with IS; (2) lack of clear understanding about their behaviour in before, during, and after the impact of the disruptive forces; (3) the long horizon of risk management that covers the entire lifespan of the system [7, 8]. This Ph.D. project aims to address these three challenges by: (1) combining artificial intelligence techniques with probabilistic approaches to model the uncertainties of these systems; (2) developing predictive simulation models to forecast the behaviour of the system; and (3) developing adaptive decision-making framework to address the long horizon of decision-making. The potential candidate will acquire knowledge about risk management practices, climate induced risks to IS, AI techniques for uncertainty modelling, agent-based modelling and system dynamics simulation, and multi-criteria decision-making.