Recent studies increasingly adopt simulation-based machine learning (ML) models to analyse critical infrastructure system resilience. For realistic applications, these ML models consider the component-level characteristics that influence the network response during emergencies. However, such an approach could result in a large number of features and cause ML models to suffer from the ’curse of dimensionality’. A clustering-based method is presented that simultaneously minimises the problem of high-dimensionality and improves the prediction accuracy of ML models developed for resilience analysis in large-scale interdependent infrastructure networks. The methodology has three parts: (a) generation of simulation dataset, (b) network component clustering, and (c) dimensionality reduction and development of prediction models. First, an interdependent infrastructure simulation model simulates the network-wide consequences of various disruptive events. The component-level features are extracted from the simulated data. Next, clustering algorithms are used to derive the cluster-level features by grouping component-level features based on their topological and functional characteristics. Finally, ML algorithms are used to develop models that predict the network-wide impacts of disruptive events using the cluster-level features. The applicability of the method is demonstrated using an interdependent power-water-transport testbed. The proposed method can be used to develop decision-support tools for post-disaster recovery of infrastructure networks.@en

In this study, we propose a methodological framework to identify and evaluate cost-effective pathways for enhancing resilience in large-scale interdependent infrastructure systems, considering decision-makers’ risk preferences. We focus on understanding how decision-makers with varying risk preferences perceive the benefits from infrastructure resilience investments and compare them with upfront costs in the context of high-impact low-probability (HILP) events. First, we compute the costs of interventions as the sum of their capital costs and maintenance costs. The benefits of the interventions include the reduction in physical damage costs and business disruption losses resulting from the improved resilience of the network. In the final stage, we develop statistical models to predict the perceived net benefits of different network resilience configurations in power, water, and transport networks. These models are employed in an optimization framework to identify optimal resilience investment pathways. By incorporating Cumulative Prospect Theory (CPT) in the optimization framework, we show that decision-makers who assign higher weights to low probability events tend to allocate more resources towards post-disaster recovery strategies leading to increased resilience against HILP events, like earthquakes. We illustrate the methodology using a case study of the interdependent infrastructure network in Shelby County, Tennessee.

@en

Enhancing the resilience of critical infrastructure systems requires substantial investment and entails trade-offs between environmental and economic benefits. To this aim, we propose a methodological framework that combines resilience and economic analyses and assesses the economic viability of alternative resilience designs for a Water Distribution System (WDS) and its interdependent power and transportation systems. Flow-based network models simulate the interdependent infrastructure systems and Global Resilience Analysis (GRA) quantifies three resilience metrics under various disruption scenarios. The economic analysis monetizes the three metrics and compares two resilience strategies involving the installation of remotely controlled shutoff valves. Using the Micropolis synthetic interdependent water-transportation network as an example, we demonstrate how our framework can guide infrastructure stakeholders and utility operators in measuring the value of resilience investments. Overall, our approach highlights the importance of economic analysis in designing resilient infrastructure systems.@en

Extreme weather events may disrupt port and coastal freight operations, resulting in direct and indirect economic losses to ports, supporting infrastructure, and reliant industry systems. Therefore, understanding the existing resilience capacity of the Texas port system to extreme weather events is necessary. To accomplish this, the weather hazards present along the Texas Gulf Coast are quantified. Stakeholder workshops, surveys, and interviews inform our understanding of the current status of resilience practice in Texas ports. Frameworks for assessing the criticality, vulnerability, exposure, risk, and resilience of the Texas port system are developed. The economic impacts of port disruptions from hurricanes are quantified using input-output tables. A tool, PortRESECO, is developed to be used by port stakeholders to assess the resilience of a port facility and view the results of the economic analysis. Finally, recommendations are made for implementation to increase the resilience of coastal freight operations in Texas.@en

Integrated simulation models are emerging as an alternative for analyzing large-scale interdependent infrastructure networks due to their modeling advantages over traditional interdependency models. This paper presents an open-source integrated simulation package for the component-level analysis of interdependent power-, water-, transport networks. The simulation platform, named ’InfraRisk’ and developed in Python, can simulate network-wide effects of disaster-induced infrastructure failures and subsequent post-disaster restoration. InfraRisk consists of an infrastructure module, a hazard module, a recovery module, a simulation module, and a resilience quantification module. The infrastructure module integrates existing infrastructure network packages (wntr for water distribution systems, pandapower for power systems, and a static traffic assignment model for road transport systems) through an interface that facilitates the network-level simulation of infrastructure failures. The hazard module generates infrastructure component failures based on various disaster characteristics. The recovery module determines repair sequences and assigns repair crews based on predefined heuristics-based recovery strategies or model predictive control (MPC) based optimization. Based on the schedule, the simulation module simulates the consequences of the disaster impacts and the recovery actions on the performance of the interdependent network. The resilience quantification module offers system-level and consumer-level metrics to quantify both the risks and resilience of the integrated infrastructure networks against disaster events. InfraRisk provides a virtual platform for decision-makers to experiment and develop region-specific pre-disaster and post-disaster policies to enhance the overall resilience of interdependent urban infrastructure networks.@en

Recent studies increasingly adopt simulation-based machine learning (ML) models to analyze critical infrastructure system resilience. For realistic applications, these ML models consider the component-level characteristics that influence the network response during emergencies. However, such an approach could result in a large number of features and cause ML models to suffer from the `curse of dimensionality'. We present a clustering-based method that simultaneously minimizes the problem of high-dimensionality and improves the prediction accuracy of ML models developed for resilience analysis in large-scale interdependent infrastructure networks. The methodology has three parts: (a) generation of simulation dataset, (b) network component clustering, and (c) dimensionality reduction and development of prediction models. First, an interdependent infrastructure simulation model simulates the network-wide consequences of various disruptive events. The component-level features are extracted from the simulated data. Next, clustering algorithms are used to derive the cluster-level features by grouping component-level features based on their topological and functional characteristics. Finally, ML algorithms are used to develop models that predict the network-wide impacts of disruptive events using the cluster-level features. The applicability of the method is demonstrated using an interdependent power-water-transport testbed. The proposed method can be used to develop decision-support tools for post-disaster recovery of infrastructure networks.@en

Among natural disasters, hurricanes pose significant threat to port infrastructure in the United States, especially to those along the coasts of the Atlantic Ocean and the Gulf of Mexico. The operational continuity of ports is critical because of the growing reliance of domestic and international businesses on ports and their significance in the growth of national and regional economies. While direct physical damages to ports are not rare, a shutdown of port operations is more likely declared by authorities as a precautionary and preparatory measure during hurricanes. Considering the enormous economic importance of ports and their vulnerability to hurricanes, the current study proposes a framework and methodology to analyze the economic risks of such hurricane-related shutdowns using hurricane- and port-related determinants. The risks of shutdowns are modeled using regression analysis based on historical port shutdown data and are combined with extensions of the well-known input–output model to predict the operational and economic risks of ports to hurricanes. The application of the methodology is demonstrated by conducting a case study based on the Texas Port System to evaluate the economic risks of hurricane-related port disruptions on the U.S. economy.@en

1. Modern infrastructure systems are highly interconnected and comprised of geographically extensive networks (Chang, 2016). Such interconnections lead to interdependencies among infrastructure sys...@en

Climate change is increasing the frequency and the intensity of weather events, leading to large-scale disruptions to critical infrastructure systems. The high level of interdependence among these systems further aggravates the extent of disruptions. To mitigate these impacts, models and methods are needed to support rapid decision-making for optimal resource allocation in the aftermath of a disruption and to substantiate investment decisions for the structural reconfiguration of these systems. In this paper, we leverage infrastructure simulation models and Machine Learning (ML) algorithms to develop resilience prediction models. First, we employ an interdependent infrastructure simulation model to generate infrastructure disruption and recovery scenarios and compute the resilience value for each scenario. The infrastructure-, disruption-, and recovery-related attributes are recorded for each scenario and ML algorithms are employed on the synthetic dataset to develop accurate resilience prediction models. The results of the prediction models are analyzed and possible design strategies suggested based on the resilience enhancement attributes. The proposed methodology can support infrastructure agencies in the resource-allocation process for pre- and post-disaster interventions.@en

Multiple criteria decision analysis (MCDA) is a decision support tool widely used by government agencies for evaluating, assessing, and prioritizing project alternatives in circumstances where conflicting and competing objectives are to be achieved. MCDA techniques offer a systematic procedure for treating a complex project selection problem into a group of simpler subproblems for identifying the best project alternative from an available pool of options while ensuring that the concerns of stakeholders are adequately addressed. MCDA is implemented in several stages, focusing on identification of project alternatives, definition of relevant criteria, assessment of performance of alternatives based on those objectives, and identification of the best alternative(s). There are several techniques available for implementing MCDA; however, each of these techniques varies in the manner stakeholder inputs are gathered and handled. Hence, the selection of a method must be based on a comprehensive analysis of the advantages and limitations of available MCDA techniques.@en

In infrastructure networks, each of the constituent infrastructure components (nodes and links) has its own significance. The extent to which each node is important to the network largely varies based on the resource or service it provides and the type of infrastructure nodes that are dependent on it. Similarly, the links are essential for the uninterrupted supply of resources to their respective destinations. To enhance the resilience of infrastructure networks, it is of vital importance to identify and quantify the consequences of infrastructure component failures on other components in the network, for which researchers have proposed several methods. Though all infrastructure nodes may be vulnerable to some kind of external hazards, resource and time constraints make it impossible for infrastructure managers and policymakers to adopt measures to enhance the resilience of every component in a network. Hence, it is necessary to address the issue of node and link prioritization to develop effective resilience strategies for networks. This study proposes two resilience indexes, a node criticality index and a node susceptibility index, to deal with two types of node ranking relevant to infrastructure network resilience. Then the node ranking indexes are used in a heuristic algorithm to rank and prioritize infrastructure links. The indexes can support decision-making for designing and managing resilience in interdependent infrastructure networks prior to a disaster, for identifying infrastructure components that warrant immediate restoration during or after a disaster, and for devising additional resilience strategies to handle heightened disaster risks during recovery.@en

Purpose: Resource allocation is essential to infrastructure management. The purpose of this study is to develop a methodological framework for resource allocation that takes interdependencies among infrastructure systems into consideration to minimize the overall impact of infrastructure network disruptions due to extreme events. Design/methodology/approach: Taking advantage of agent-based modeling techniques, the proposed methodology estimates the interdependent effects of a given infrastructure failure which are then used to optimize resource allocation such that the network-level resilience is maximized. Findings: The findings of the study show that allocating resources with the proposed methodology, where optimal infrastructure reinforcement interventions are implemented, can improve the resilience of infrastructure networks with respect to both direct and interdependent risks of extreme events. These findings are also verified by the results of two case studies. Practical implications: As the two case studies have shown, the proposed methodological framework can be applied to the resource allocation process in asset management practices. Social implications: The proposed methodology improves the resilience of the infrastructure network, which can alleviate the social and economic impact of extreme events on communities. Originality/value: Capitalizing on the combination of agent-based modeling and simulation-based optimization techniques, this study fulfills a critical gap in infrastructure asset management by incorporating infrastructure interdependence and resilience concepts into the resource allocation process.@en

Over the past years, the frequency and scope of disasters affecting the United States have significantly increased. Government agencies have made efforts in improving the nation's disaster response framework to minimize fatalities and economic loss due to disasters. Disaster response has evolved with the emergency management agencies incorporating systematic changes in their organization and emergency response functions to accommodate lessons learned from past disaster events. Technological advancements in disaster response have also improved the agencies' ability to prepare for and respond to natural hazards. The transportation and logistics sector has a primary role in emergency response during and after disasters. In this light, this paper seeks to identify how effective policy changes and new technology have aided the transportation and logistics sector in emergency response and identify gaps in current practices for further improvement. Specifically, this study compares and contrasts the transportation and logistical support to emergency relief efforts during and after two major Hurricane events in the U.S., namely Hurricane Katrina (which affected New Orleans in 2005) and Hurricane Harvey (which affected Houston in 2017). This comparison intends to outline the major steps taken by the government and the private entities in the transportation and logistics sector to facilitate emergency response and the issues faced during the process. Finally, the paper summarizes the lessons learned from both the Hurricane events and provides recommendations for further improvements in transportation and logistical support to disaster response.@en

Traffic operations vary drastically before-, during- and after natural disasters due to several reasons, such as the movement of affected populations, road failures, and heavy precipitation. The disaster-induced effects on road links and their subsequent recovery to the pre-disaster state are functions of the designed network resilience. Thus, the investigation of traffic operations including speed and volume fluctuations during such disasters can provide useful insights into the resilience of the roadway network. Furthermore, identifying road links and corridors which are most affected during natural disasters and estimating the extent of the effect on traffic are crucial steps in devising traffic management strategies for mitigating future hazards. A key challenge to the realization of the above objectives is the lack of data pertaining to roadway and traffic conditions during natural disasters. While modern sensing technologies based on non-Dedicated Short-Range (wireless) Communications such as Bluetooth® have been widely adopted to track traffic performance, they need to be corroborated with supplementary data for a complete characterization of the prevalent traffic conditions. In this study, an alternative method is presented for identifying the traffic fluctuations induced by a natural disaster (Hurricane Harvey) on an urban traffic network (Houston) by studying the characteristics of extreme travel time observations. The study relies on time series decomposition and anomaly detection algorithms to investigate the spatiotemporal effects of the hurricane on the traffic conditions. The results of the case study suggest that the metrics developed are effective in quantifying the resilience of traffic networks against natural disasters by capturing both the initial impact and recovery. The study showed that the Houston freeway network was not completely recovered from the effects of Hurricane Harvey, even three weeks after the occurrence of the hurricane.@en

Unanticipated events such as natural disasters, terrorist attacks, cyber-attacks, and so forth, could cause prolonged disruptions in major utility service networks including, for example, water and electricity, in urban areas. Owing to the presence of complex interdependencies among infrastructure systems in an urban network, the disruption of one system may trigger a chain of events that degrades the proper functioning of several other dependent systems. Consequently, many parts of the city may not have access to multiple utility services and amenities. Identifying the most vulnerable communities exposed to such utility disruptions is key to performing immediate relief operations. In this paper, the concept of Priority Index, introduced as a measure of the susceptibility of communities to the event, is presented to rank urban regions based on the extent of the impact of disruptions (both cascading and interdependent impacts) caused by an event, as well as the social vulnerability of communities. Agent-based models are employed to simulate the consequences of a disruptive event on a semi-realistic urban infrastructure network. Later, the extent of impact on communities is evaluated using the simulation results and the American Community Survey data. The proposed Priority Index could help city administrations and utility service agencies identify the regions in a city that require immediate attention after a disruptive event occurs in the infrastructure network. A case study based on a semi-realistic infrastructure network in Austin, Texas is presented to demonstrate the implementation of the concept of Priority Index and the methodological framework.@en

Free-flow speed (FFS) is the desired speed that drivers choose when no (or very less number of) vehicles are present in the road segment. FFS is an important parameter of traffic flow that decides the level of service and capacity aspects of various types of highway facilities. Estimation of FFS is extremely time consuming and requires extensive human resource and capital. Hence, a FFS model can be a solution to bring down the above difficulties while ensuring satisfactory prediction of FFS. In countries like India, a widely used method of estimating FFS is to collect vehicle speeds from field during low volume hours. However, this method requires significant amount of time, human resource and capital for studies on large road networks. Hence, it is essential to develop models to predict free-flow speeds. It is important that models are capable of capturing the free-flow speed variations due to local road factors. Majority of the existing free-flow speed models are developed under homogeneous traffic conditions, in which passenger cars dominate the vehicle composition. However, the traffic condition in emerging economies like India is heterogeneous in nature characterized by the presence of multiple vehicle categories with varying physical and dynamic characteristics. The present paper attempts to investigate the influence of different road factors on FFS on urban roads of Chennai, India. The paper also tries to capture the FFS variations across different classes of vehicles and develop FFS prediction models. The typical Indian traffic comprises significant percentage of slow moving vehicles like three-wheelers as well as fast moving sedans and SUVs. Composition of traffic and corresponding proportions of different classes are important factors that differentiate heterogeneous and homogeneous traffic. The presented models could explore the driver speed behavior with respect to the aforementioned factors into consideration.@en

Free-flow speed (FFS) is the speed of vehicles under low volume conditions, when the drivers tend to drive at their desired speed without being affected by control delay. Estimation of FFS is important in several applications. FFS varies extensively across various road facilities as they are influenced by driver behaviour, vehicle characteristics, road factors, landuse, geometric features, control factors, etc. The estimation of FFS in homogeneous traffic is comparatively simpler as the speed variation across vehicles is limited. However, in heterogeneous traffic conditions existing in countries such as India, the FFS distribution varies across vehicle classes. The studies conducted by the authors explored the FFS distribution of various vehicle classes such as two-wheelers, three-wheelers, cars, buses, etc. However, detailed analysis revealed that the variation in FFS can be better explained by further classification of vehicles into subclasses. The study also found that the vehicle's lane position is a factor affecting FFS. The study was conducted on four- and six-lane divided roads in Chennai, India. A total of 24 study sections were chosen for data collection. Speed data were collected during early morning hours to ensure free-flow conditions. The vehicle movements were recorded using video cameras. The details regarding site factors such as carriageway width, link length, landuse, presence of kerb and type of area were collected manually. The speed and lane data were extracted and tabulated from the video recordings. The authors studied the speeds of about 17,800 vehicles (36% two-wheelers, 8% three-wheelers, 8% buses, 33% cars, 10% light commercial vehicles and 5% trucks). The vehicles were classified into 14 subclasses and speeds were analysed. The study also evaluated the effect of lane position on FFS of different classes of vehicles. It was found that vehicles on kerb lanes experienced lower speeds than those on inner lanes. Furthermore, FFS models for four- and six-lane divided roads were developed using multiple linear regression. Significant difference in speeds was observed within and across subclasses of vehicles. The models also evaluated the effects of various road factors such as carriageway width, link length, adjacent landuse type and presence of kerb on FFS. Models such as these can find applications in planning and operational analysis of urban road facilities.@en