Assessing the Resilience of Critical Infrastructure Facilities toward a Holistic and Theoretical Approach: A Multi-Scenario Evidence and Case Study

Abstract

Given the increasing frequency of natural disasters, which result in substantial damage to critical infrastructures and disrupt the functioning of modern societies, numerous studies have been conducted in recent decades to propose sustainable preventive and enhancement measures to safeguard the environmental and societal development. This paper contributes to the existing literature by introducing a novel environmentally conscious infrastructural resilience assessment approach named the Novel Infrastructure Resilience Assessment Curve (NIRAC). Unlike past works which typically focused on a single infrastructure scenario, the NIRAC is conceptualized around multi-scenario resilience assets, integrating sustainable principles to enhance environmental resilience. Additionally, this paper presents a road infrastructure resilience assessment (RIRA) framework, developed from factors and dimensions pertinent to road infrastructure resilience and environmental sustainability. The RIRA framework is applied to a case study of a road damaged by heavy rains, emphasizing the need for sustainable recovery efforts which minimize environmental impact. The results demonstrate the effectiveness of the RIRA framework in guiding road recovery efforts. The insights provided in this paper are valuable for disaster managers and policymakers, as they extend their resilience assessment knowledge with a focus on sustainable development and environmental protection. This expanded knowledge facilitates the implementation of appropriate interventions to prevent and mitigate the societal consequences of disasters more efficiently.

1. Introduction

Rapid worldwide urbanization has presented a host of fundamental challenges, compounded by the threat of natural and human-made disasters, which have significant implications for environmental sustainability and the resilience of critical infrastructures. These challenges pose significant risks to the functionality of communities, cities, and other critical infrastructures, with far-reaching consequences for global economies and societies [1]. Critical infrastructures are widely recognized as vital national assets due to their role in providing essential services crucial for the sustainable development, resilient functioning, and long-term environmental and societal sustainability of communities [2]. As risks associated with these challenges continue to escalate, so too are the anticipated damages, which could result in the sustainability and functionality of infrastructure systems or even widespread failures. The ability to effectively cope with and mitigate the adverse effects of such disruptive events is known as “resilience”—a concept crucial for ensuring both environmental sustainability—and has been a longstanding challenge for engineers and decision makers from past centuries to current times [3].

Resilient infrastructures possess the capacity to adapt to changing conditions to a certain extent [4,5]. They exhibit a greater ability to anticipate and utilize diverse forms of information, as well as different methods of synchronization across multiple disasters [6,7,8]. However, if an extreme shock persists, causing changes in duration that surpass a certain intrinsic threshold of the infrastructure’s resistance level, the services provided by this infrastructure may be disrupted to such an extent that its functionality, sustainability, and environmental integrity are compromised [8,9,10,11]. Thus, it is through the balance of positive (infrastructural resistance) and negative (disaster pressure) feedback loops that an infrastructure performance level against disaster can be ascertained [11,12,13,14].

1.1. Resilience Concept and Infrastructural Resilience Assessment

The concept of resilience holds varied meanings across different fields and research purposes, often linked to sustainability, leading to diverse definitions in the literature. Originating from Holling [15] in the field of socio-ecological systems, resilience was initially described as a measure of a system’s resistance and its ability to assimilate changes and disturbances while maintaining the same relationships between populations or state variables. In the context of hazard research, resilience refers to a system’s capability to endure and survive a disaster along with minimal impact and damage on the environment and infrastructure, ensuring sustainable recovery [16]. An engineering-based perspective defined resilience as the adaptive capacity of a system, emphasizing its ability to withstand or adjust to external disturbances while maintaining environmental and infrastructural sustainability. This definition also includes the effects of pre-disaster preparedness and adaptive response measures, crucial for effective and efficient recovery processes following a disaster [17]. The concept of resilience proves challenging to define due to the multitude of definitions and approaches found in the literature [18], reflecting its broad usage across disciplines such as socio-ecology, sociology, psychology, and economics [19].

Despite its wide application, resilience has not been extensively implemented in engineering, particularly in the analysis of critical infrastructural resilience in daily practice; however, this constitutes the basis for sustainable development and the daily functioning of modern societies and economies [20]. Over the past two decades, resilience assessments have emerged as a prominent topic, quantifying a system’s ability to prepare, plan, recover, and adapt to adverse events more successfully. Assessing the resilience of critical infrastructures has become a specific, professional process, requiring adherence to basic principles such as complexity, adequacy, specificity, expertise, and impartiality [21]. Consequently, resilience assessments of infrastructure must provide solutions regarding infrastructure system preparedness, the capacity to absorb consequences, effective response, and adaptation to new conditions.

Resilience assessments play a crucial role in ensuring the safety and environmental sustainability of not only individual components within a system but also the entire infrastructure system [1,22]. In this context, authors have developed resilience assessment models that statistically evaluate the resilience level of critical infrastructural elements, requiring a comprehensive assessment of their robustness, recoverability post disaster, and ability to adapt to previous adverse events. Recently, a different conceptualization of resilience assessment was proposed by Ouyang [23], introducing a three-stage concept (i.e., disaster prevention, damage propagation, and recovery) for infrastructural resilience assessments. They highlighted the continuous evolution of infrastructure systems alongside hazard occurrence rates and intensities, emphasizing the need for sustainable adaptation strategies. As a result, they suggested a time-dependent resilience metric, given by the ratio between the time integral of the actual performance curve and the time integral of the target performance curve. These integrals are constrained by the lifetime of an infrastructure system, allowing for multiple events to occur, as well as future system and environmental evolution [24].

Resilience is a multidisciplinary term that finds utility across various sectors, each with its own set of dimensions. Among these dimensions, the most frequently cited are robustness, redundancy, resourcefulness, and recovery, collectively known as the 4Rs [6]. These dimensions are widely applicable across industries and are fundamental to understanding resilience. To obtain evidence related to the transportation sector, there has been a significant focus on studying resilience since 2009 from different research perspectives, including roadways and highways [11,16]. Researchers have offered diverse definitions of resilience and have utilized various dimensions to measure the resilience of transportation infrastructure. The dimensions often used in the literature are absorptive, adaptive, and restorative capacity. Many researchers have expanded upon these dimensions to measure the resilience of transportation systems further, including factors such as efficiency, diversity, performance, and mobility [9,19]. Researchers have developed several dimensions to measure the resilience of transport infrastructures while considering the interdependence between different critical infrastructures. However, only a few of them are fully capable of measuring the resilience of an existing road as a standalone infrastructure [6,25].

1.2. Curve-Based Conceptualization for Infrastructure Resilience Assessment

The integration of the resilience concept in various scientific fields has garnered significant interest in the existing literature, especially with the increasing focus on sustainability. While the engineering domain is still in its early years, there is a considerable number of studies that incorporate resilience into the assessment and protection of critical infrastructure vulnerability, emphasizing the need for sustainable and environmentally sound practices. Recent studies have focused on evaluating infrastructure resilience by considering performance levels before, during, and after a disaster event, including infrastructures’ ability to recover and return to normal functionality. Curve-based resilience assessment methods have been widely used in the literature as a tool to quantitatively and qualitatively communicate aspects of infrastructural performance to stakeholders [25].

The functionality of an infrastructure is typically normal during the pre-disaster period, meaning that it operates and provides services as usual, contributing to the sustainability of community services and environmental systems. However, when a disaster event occurs, the infrastructure may lose its normal state, and the level of service it can provide will depend on its ability to withstand the recognized shocks. After the event, the infrastructure requires intervention for recovery. The vulnerability of the infrastructure plays a crucial role in determining the time and resources needed for its recovery. Infrastructures with a lower vulnerability will require a quicker recovery time with limited resources to return to their initial function. Conversely, infrastructures with higher vulnerability levels may require more time and resources for recovery. This relationship between vulnerability and recovery time is illustrated in Figure 1. The time it takes for an infrastructure facility to return to its initial function with limited resources, along with its ability to adapt after recovery, significantly influences its resilience [26,27,28]. Resilience, in this context, is closely linked to the infrastructure’s ability to recover efficiently and effectively from disruptive events, ensuring a timely restoration of services to meet societal needs.

1.3. Background on Road Infrastructure Resilience Assessment

Resilience is a crucial component for critical infrastructure facilities to maintain functionality during disruptive events [30]. Resilient road infrastructure has the ability to reduce the impact intensity of crises, reduce failure duration, and speed up the recovery time [31]. Given the substantial contribution of road infrastructure to the development and well-being of modern society [32,33], assessing its performance when faced with disastrous events becomes a critical step toward enhancing its resilience and environmental sustainability. While few studies can comprehensively measure the resilience of an existing roadway as a standalone infrastructure, the existing literature has highlighted numerous dimensions of road infrastructure that must be considered when evaluating its resilience. Several dimensions have been proposed to assess the resilience of road infrastructure. Nipa et al. [34] suggest considering the number of intersections (nodes), length of the roadway without nodes (link), total length of disrupted roadway, existence of optional routes, availability of budget, and preparedness actions. Aghababaei et al. [35] have introduced a new measure called trip resilience (TR) to evaluate the resilience of travel on road networks post disaster, integrating the dimensions of robustness, redundancy, and recovery. Nipa, Kermanshachi et al. [36] have listed dimensions like the presence of a railroad crossing, distance of the link/node from the affected area, level of road damage, availability and access to previous disaster data for the roadway, regular funding for resilience enhancement activities, etc., as key factors for assessing road infrastructure resilience and sustainability.

The literature has largely overlooked the ability of road infrastructure to detect danger and transfer information as a critical dimension of road resilience. This paper emphasizes the intelligence ability of roads for information dissemination as an essential aspect of road infrastructure resilience assessments. Evaluating the intelligence ability of road infrastructure before a disaster allows policymakers to take effective actions for disaster preparedness and the mitigation of consequences. The second dimension considered in this research is the robustness of road infrastructure. Robustness refers to the infrastructure’s ability to be strong enough to withstand and resist event strikes, ensuring its sustainability during adverse conditions [36]. The third considered dimension is the recoverability of the road, which is represented by the speed of time which disrupted road infrastructure will require to retake normal function, contributing to a swift and sustainable recovery [37,38].

Multiple resilience infrastructure assessment frameworks exist in the literature. Thus, Vugrin et al., in 2011 [39], developed a comprehensive resilience assessment framework for evaluating the sustainability and resilience of infrastructure systems. This framework employs both quantitative and qualitative methodologies to measure resilience resulting from disruptions to infrastructure function. It also analyzes system characteristics affecting resilience to provide insights and directions for potential improvements. From the acknowledged lack of an integrated framework which takes into account the nature and sequence of multiple hazards and their impact, different restoration strategies, and, therefore, the quantification of resilience in the literature, Argyroudis [1,38] proposed for the first time a new framework for quantitatively assessing the resilience of critical infrastructures exposed to multiple hazards, taking into account the vulnerability of assets to hazardous actions and the speed of damage recovery, including temporal variability in the hazards. Liu et al. [40] created an exploratory resilience assessment framework of an urban road network that includes the resilience performance index, the robustness index, and the recovery index for sustainable cities.

With regard to multiple disruptive events such as floods, earthquakes, etc., which result in significant damage to critical infrastructures, it is essential to propose and define preventive actions and measures that can be implemented to mitigate their consequences. In response to this need, a Civil Infrastructure Resilience Assessment Framework (CIRAF) was introduced by Singhal et al. [41] to assess the seismic fragility and resilience of single or multiple interconnected civil infrastructure systems following a disruptive event. To respond to and handle the impact of these disasters on critical infrastructure, Guo [42] reviewed 24 resilience assessment frameworks for critical infrastructures, provided by 24 quality papers published over the past decade. They determined and analyzed the common dimensions together with the key indicators of resilience assessment frameworks and proposed possible opportunities for future research.

Assessing the performance level of an infrastructure system exposed to hazardous events enables decision makers and managers to determine appropriate operational measures for its enhancement, ensuring the sustainability and environmental resilience of critical systems [10]. Based on the fundamental infrastructure resilience assessment curve adapted by Bi [20], the objectives of this paper are the following: (i) We wish to develop and propose a new infrastructure resilience assessment model, grounded in sustainability principles and based on existing conceptualization curves in the literature. This model is named the Novel Infrastructure Resilience Assessment Curve NIRAC. The key distinction of the NIRAC from previous infrastructure resilience assessment curves is that it is conceptualized based on multiple scenarios of infrastructure, considering sustainability and environmental impact, rather than a single scenario as in past works. Resilience infrastructure assessments using NIRAC span the lifetime of a disaster. (ii) We also aim to formulate and provide a road infrastructure resilience assessment (RIRA) framework by identifying relevant road resilience dimensions and their corresponding resilience factors. (iii) We will also explore the inter-relationship between road infrastructure resilience factors within the performance curve, highlighting their role in promoting sustainability and reducing environmental impact. (iv) Finally, we wish to apply the proposed RIRA framework in a real road case study to verify its feasibility, ensuring that the framework can support sustainable infrastructure development. The selected case project is the National Road no.9 (NR9) in Burundi, which is usually significantly damaged by landslides caused by heavy rainfall.

The remainder of this paper is structured as follow: After this Introduction (Section 1), Section 2 follows, which provides this research paper’s Materials and Methods. Section 3 presents the results of this study by introducing the Novel Infrastructure Resilience Assessment Curve (NIRAC) and detailing the development and presentation of the road infrastructure resilience assessment (RIRA) framework. Section 3 also presents the empirical implementation of the RIRA framework in a critical road case study, along with discussions and recommendations related to the case study. Section 4 finally concludes this research paper, underscoring the importance of sustainability and environmental resilience in infrastructure resilience assessments.

2. Materials and Methods

The methodology we used to reach our research goal consisted of the following steps:

• First, we conducted a thorough review of the concept of resilience, with a focus on sustainability and the methodologies for assessing critical infrastructure resilience. The following keywords were utilized to search peer-reviewed publications: infrastructure resilience, sustainable infrastructure systems, critical infrastructure, road infrastructure resilience, resilience assessment, and critical infrastructure resilience assessment. These keywords were searched across databases including Web of Science, Scopus, and Google Scholar. A total of 307 research articles, 89 conference papers, and 52 books published between 2010 and 2023 were reviewed. This critical review aimed to understand the current needs and techniques for infrastructure resilience assessment, highlighting gaps in the application of infrastructure resilience, as understanding these existing deficiencies would enable the identification of pathways to contribute to and address the current gaps in the field. Based on the aim of this research, the literature enabled us to further identify critical infrastructure resilience dimensions and their corresponding sustainability factors specifically for roadway infrastructure, allowing us to develop our road infrastructure resilience assessment framework. The distribution of the reviewed publications is depicted in Figure 2.

• We applied a system thinking approach to identify the dimensions and factors that influence road infrastructure resilience, with an emphasis on sustainability. System thinking, which fosters curiosity, humility, and openness to multiple perspectives and possibilities, was applied in a holistic manner to analyze resilience within its environmental context. This approach enabled us to identify and address the fundamental factors of resilience, ensuring that they were aligned with sustainable objectives. Based on this approach, the authors adopted “robustness”—which refers to the ability of an infrastructure to withstand and cope with the occurrence of disasters and resist any negative impacts caused by such events—as a crucial dimension for a resilient road as a standalone infrastructure, as it contributes to the strengthening of the road infrastructure. A resilient road must be robust enough to ensure that its pavement can physically withstand disruptive events. Modern infrastructures must maximize the comfort and safety of users, making it essential for the road infrastructure to be intelligent enough to monitor and control daily climate change and traffic conditions, providing data through an equipped information dissemination system. Thus, the “Intelligence” dimension contributes significantly to the resilience of road infrastructure. “Recoverability” is the ability of road infrastructure to quickly resume operation after being damaged or destroyed.
• The identified factors enabled us to develop and formulate a road infrastructure resilience assessment framework, which was then applied to a case study of a road damaged by a natural disaster in Burundi. The research methods and procedures for this study are detailed in Figure 3.

3. Results

3.1. Novel Infrastructure Resilience Assessment Curve (NIRAC)

Infrastructure resilience assessment curves have been a frequently used approach among the various resilience assessment tools, playing a crucial role in sustainable infrastructure planning. First introduced by Bruneau [29] to quantitatively present a conceptual model to define the seismic resilience of communities, this approach has since gained great importance across various research areas. However, previous research has typically presented a performance curve for a single system scenario to assess infrastructure performance. This study proposes a more proficient and skillful approach to evaluating infrastructure resilience by comparing the performance curves of multiple scenarios of an infrastructure exposed to disaster events.

The four scenarios (A, B, C, and D) describe, respectively, the degree of disruption (severe, large, minor, and slight/no) for infrastructures under disaster conditions. In Scenario A, severe disruption occurs with widespread damage or failure across the infrastructure networks. Scenario B models a larger-scale disruption, such as a widespread power failure or major disturbances across critical infrastructure systems. In contrast, Scenario C introduces a minor disruption, where small-scale system failures occur, but the overall impact on infrastructure performance remains limited. Finally, Scenario D represents a situation where an infrastructure operates under normal or slightly stressed conditions, without any significant disruptions. The four scenarios offer a clear, structured analysis of infrastructure resilience, progressively representing different levels of vulnerability. This division allows for a comprehensive understanding of how various infrastructures respond, resist, and recover from disaster events.

Furthermore, the NIRAC uses identical ranges across the four scenarios for ensuring a consistent basis for their comparison, and this decision was made to eliminate any variability that may be caused by differing ranges. This consistency enables for a controlled assessment of how varying situations such as different inputs can affect the outcomes. It eliminates confounding variables and allows the results to be assigned merely to the key differences between scenarios, instead than variations in ranges. Figure 4 illustrates the performance of four scenarios (A, B, C, and D) of an infrastructure facility in resilience assessment curves subjected to disaster events.

NIRAC Explanation

Here is a detailed explanation based on the provided description of the infrastructure resilience assessment curves for scenarios A, B, C, and D:

• In the pre-disaster period, the infrastructure operates at its normal performance level, providing maximum service, as expected. During this time, the degree of vulnerability of the system is determined by the threshold of a span angle of 180 degrees. This indicates that the infrastructure has not experienced any disruptions, and its robustness is still undetermined. Therefore, all four considered infrastructure scenarios (A, B, C, and D) are providing their expected services without any issues or disruptions.
• During the occurrence of the event, the performance of the infrastructure in four scenarios (A, B, C, and D) varies based on their behavior against the disaster. Each scenario’s vulnerability level is determined by its state of disruption, along with its level of resistance to providing service. The difference in the vulnerability level is determined by the gap in the vulnerability degree. Therefore, the infrastructure scenario with a direct loss of function due to the occurrence of the events seems more vulnerable and less robust, so the angle between the infrastructure’s initial function and vulnerability level is estimated to be 90° and the climbing vulnerability level is estimated to measure n = 15° from a scenario to another.
  Note that n = 15° is a random escalation degree between one scenario and another that can change according to the value of the vulnerability level of the scenarios.

3.

The four presented scenarios are deescalated by n = 15°. This means that, if the infrastructure presents a performance which is 15° more than that of another, then it is more robust. Furthermore, the infrastructure in Scenario A, with a vulnerability level of 90°, is less robust than B, measuring about 15° more. So, B’s robustness level threshold is equal to 90° + 15° (90° + n). However, infrastructure Scenario B, with 90° + n, is less robust than infrastructure C, with a resistance curve of 90° + 2n, which is, in turn, less robust than infrastructure system D, whose level of resistance is determined by a span angle of 90° + 3n.

4.

Based on the general definition of infrastructure resilience, which is the ability of an infrastructure to withstand, respond, and recover from the effect of a disruptive event [2,25], this ability is described in function of time and the performance level of the targeted infrastructure. Therefore, a resilience assessment curve, as illustrated in Figure 4 for the recovery aspect, presents a different performance from one scenario to another. The assessment method proves that, the more an infrastructure is robust, the more it will be rapidly repaired or recovered. For this reason, the time between ta and ta’ taken by infrastructure Scenario A (with 90°) to be recovered is too long and requires more resources than the time tb − tb’ taken by infrastructure Scenario B (with 90° + n) to be recovered, and so on.

Table 1. Performance level of scenarios and relative impact during the disaster lifespan.

Scenario	Pre-Disaster Performance	Performance During the Disaster	After-Disaster Status
Scenario A	provides full service	the most vulnerable	high vulnerability, so it requires significant time and resources to be recovered
Scenario B	provides full service	less disruption	requires fewer resources and less time for restoration
Scenario C	provides full service	greater resilience and more robust	requires a shorter recovery time
Scenario D	provides full service	the most resilient and the least disruption	the fastest, the least required resources, and the most robust

summarizes the performance level of each scenario during the disaster and the corresponding impact during the recovery process for each scenario studied.

3.2. Road Infrastructure Resilience Assessment (RIRA) Framework Development

3.2.1. System Thinking Approach for Road Infrastructure Resilience Factor Identification

System thinking is a holistic way to bring sustainable solutions to more effectively solve a problem by addressing the root factors and causes of an issue whilst considering the complex system in which it appears. It was found to be a suitable and helpful approach to identify road resilience dimensions and factors according to their influence capacity in terms of road disaster information dissemination, road disaster resistance and response, and rapid road recovery. Figure 5 illustrates and defines in detail the system thinking approach that needs to be followed to identify the resilience of road infrastructure dimensions, based on this research, forming the RIRA framework.

This study’s secondary objective was to assess the resilience levels of roads and highway networks, thus focusing on identifying the most relevant dimensions for these systems. Following the establishment of road infrastructure dimensions, the research progressed, identifying specific road resilience factors corresponding to each dimension. These factors were chosen for their close alignment with road facilities and could be easily measured. After having been identified by the authors, the factor list was sent to experts for validation.

The initial list of the identified factors for road resilience was composed of sixty-eight factors. However, after a rigorous review by experts, the final list was reduced to fifty-two factors. Some factors were deleted due to their inadequacy for standalone road infrastructure, while others were combined due to their similar meaning and influence on road infrastructure resilience. For example, factor “Traffic origin and destination monitoring” initially identified by the authors was renamed by the experts to “Traffic route control”. Factors “Roadside vegetation” and “Availability of Planting zones”, which had been separately identified by the authors, were combined and renamed “Roadside vegetation condition and planting” by the experts. Due to the similarity of factors such as “Concrete strength”, “Surface porosity”, “Electromagnetic compatibility”, etc., with others identified by the authors, the experts suggested that these factors be deleted, a suggestion which was consequently considered by the authors.

Table 2. Road infrastructure intelligence factors.

Dimensions	Indicators	Factors	ID
Road Intelligence	Weather condition monitoring	Wind speed and direction monitoring	IN-11
		Rain, storm, snow, fog, and ice monitoring	IN-12
		High-temperature monitoring	IN-13
		Thunder and lightning monitoring	IN-14
	Traffic condition monitoring	Traffic volume and flow congestion monitoring	IN-21
		Traffic crashes/collision monitoring	IN-22
		Traffic route control	IN-23
	Road facility conditions	Pavement temperature and humidity monitoring	IN-31
		Road lighting and marking	IN-32
		Road shape monitoring (slope, curve, etc.)	IN-33
		Road intersections system	IN-34
	Vehicle conditions	Vehicle load and motion monitoring	IN-41
		Vehicle mass and noise control	IN-42
		Vehicle size and speed monitoring	IN-43

,

Table 3. Road infrastructure robustness factors.

Dimension	Indicators	Factors	ID
Road Robustness	Road Pavement Conditions	Soil type conditions (gravel, sand, clay, and silt)	RO-11
		Soil property (stiffness or elasticity of the soil)	RO-12
		Temperature and moisture content	RO-13
		Roadbed strength and durability	RO-14
		Degree of compaction/subgrade stabilization	RO-15
		Construction and material quality	RO-16
		Surface permeability (void rate) and friction	RO-17
		Corrosion and drought resistance	RO-18
		Adhesion degree of aggregate with bitumen	RO-19
		Skid resistance of surface texture	RO-110
	Roadside and Protection facilities	Side slope and edge protection	RO-21
		Dust removal system	RO-22
		Presence of guardrail	RO-23
		Availability of sidewalk zones	RO-24
		Roadside vegetation conditions and planting	RO-25
		Ditch and road embarkment conditions	RO-26
	Road Drainage System	Roadside and surface drainage	RO-31
		Cross- and sub-surface drainage	RO-32
	Road Signage System	Road marking condition	RO-41
		Road lighting system	RO-42
		Installation of RFID system	RO-43
		Road signs and symbols	RO-44

and

Table 4. Road infrastructure recoverability factors.

Dimension	Indicators	Factors	ID
Road Recoverability	Resource-related factors	Availability of fund/budget	RE-11
		Skilled and experienced workforce	RE-12
		Availability of construction materials	RE-13
		Equipment availability and workability	RE-14
	Social factors	Local maintenance feasibility	RE-21
		Part-time inspection	RE-22
		Reporting mechanism	RE-23
		Land acquisition management	RE-24
	Technical factors	Rapid design	RE-31
		Prefabricated construction materials	RE-32
		Traffic diverging	RE-33
		Alternative route	RE-34
	Organizational factors	Road recovery processes	RE-41
		Quick decision mechanism	RE-42
		Special department establishment	RE-43
		Key stakeholders’ involvement and commitment	RE-44

present a final list of the identified factors influencing road infrastructure resilience, classified according to the three related dimensions of standalone road infrastructure.

3.2.2. Road Resilience Factor Analysis

The identified factors were shown to experts in disaster management and post-disaster road reconstruction for evaluation, weighting, and validation. The experts were asked to score each factor using a Likert scale ranging from 1 to 5, where 1 signified “Most important” and 5 “Least important”. The collected data were analyzed using descriptive statistical methods with the IBM SPSS V24 software to determine the mean score and significance of each factor. The respondents (experts) provided their opinions by rating each factor according to its importance. The objective of this pilot survey was to evaluate the significance of these factors to the resilience of road infrastructure and validate them based on the perception of the expert respondents.

Table 5. Mean score and significance status of road resilience factors.

Road Intelligence-Related Factors				Road Robustness-Related Factors				Road Recoverability-Related Factors
Rank	Mean	SD	p-Value	Rank	Mean	SD	p-Value	Rank	Mean	SD	p-Value
IN-12	4.83	0.391	0.000 a	RO-25	4.48	0.571	0.001 a	RE-13	4.51	0.418	0.002 a
IN-11	4.69	0.400	0.001 a	RO-41	4.41	0.585	0.001 a	RE-14	4.36	0.424	0.004 a
IN-32	4.68	0.402	0.001 a	RO-31	4.36	0.594	0.002 a	RE-21	4.24	0.456	0.007 a
IN-13	4.62	0.406	0.002 a	RO-14	4.34	0.597	0.003 a	RE-33	4.20	0.465	0.008 a
IN-22	4.60	0.408	0.002 a	RO-21	4.32	0.600	0.003 a	RE-23	4.12	0.473	0.011 a
IN-41	4.55	0.412	0.002 a	RO-26	4.23	0.611	0.003 a	RE-32	4.12	0.473	0.011 a
IN-21	4.36	0.436	0.004 a	RO-23	4.17	0.620	0.004 a	RE-22	3.99	0.486	0.019 a
IN-23	4.33	0.444	0.005 a	RO-17	4.12	0.633	0.005 a	RE-31	3.97	0.492	0.020 a
IN-33	4.29	0.452	0.005 a	RO-43	4.02	0.648	0.007 a	RE-24	3.92	0.499	0.022 a
IN-14	4.23	0.463	0.006 a	RO-11	4.00	0.651	0.008 a	RE-12	3.86	0.506	0.025 a
IN-31	4.12	0.480	0.008 a	RO-44	3.96	0.656	0.009 a	RE-34	3.86	0.506	0.025 a
IN-43	4.00	0.491	0.009 a	RO-24	3.93	0.660	0.010 a	RE-42	3.84	0.510	0.030 a
IN-34	3.99	0.498	0.022 a	RO-110	3.92	0.661	0.010 a	RE-11	3.78	0.521	0.031 a
IN-42	3.97	0.502	0.023 a	RO-32	3.91	0.667	0.010 a	RE-43	3.76	0.532	0.036 a
				RO-12	3.87	0.674	0.011 a	RE-41	3.68	0.558	0.041 a
				RO-18	3.86	0.678	0.011 a
				RO-19	3.79	0.682	0.016 a
				RO-13	3.77	0.686	0.017 a
				RO-16	3.74	0.695	0.023 a
				RO-15	3.71	0.700	0.031 a
				RO-42	3.68	0.704	0.033 a
				RO-22	3.66	0.707	0.041 a

^a Weighting result of the factor was significant at the 0.05 level.

presents the final results of the analysis, including the mean score and significance level of each road resilience factor.

According to the theory presented by [37], the reasoning low mean score is non-significant at 3.00 and indicated by a p-value ≥ 0.05. Based on the results presented in Table 4, the identified factors were found to have high mean scores and significance levels according to the experts’ point of view. Then, these factors were all considered to assess the resilience of road infrastructure.

3.2.3. Inter-Relationship between Road Infrastructure Resilience Factors within the Performance Curve

The performance level of road infrastructure before the occurrence of a disaster is determined by the intelligence dimension factors. Thus, road infrastructure will provide the intended service as long as it is capable of monitoring weather and road traffic conditions, road facilities, and vehicle conditions. During a disaster event, the performance of road infrastructure facilities will be determined by the road base, surface, and side conditions, the resistance of its protection facilities, and drainage and signal system conditions. After a disaster, road infrastructure proceeds with its recovery processes. During this phase, factors related to resource availability and allocation influence the speed of recovery. Furthermore, managing social, technical, and organizational factors contributes greatly to an effective and efficient recovery process. Figure 6 illustrates the performance curve of a road infrastructure subjected to a natural disaster event and classifies resilience factors based on the phase of the disaster lifecycle.

3.2.4. Road Infrastructure Resilience Assessment (RIRA) Framework

Resilience assessment plays a crucial role in ensuring the sustainability and security of critical infrastructure in the face of natural disasters. To enhance the existing resilience assessment frameworks highlighted in Section 1.3, this paper provides a novel framework, specifically tailored to road infrastructure assets, a topic which has been overlooked by researchers in this field. The proposed framework, called the road infrastructure resilience assessment (RIRA) framework, was formulated based on distinct dimensions of road infrastructure resilience and their related factors. Within this framework, three key dimensions of road resilience (intelligence, robustness, and recoverability) are identified, along with their respective indicators and factors, forming the foundational structure for the RIRA framework. This framework guides the assessment of road resilience throughout the entirety of a disaster’s lifecycle. Figure 7 illustrates the RIRA framework proposed in this paper.

Our model was specifically developed to address the limitations of existing road resilience assessment models, particularly in accounting for real-time dynamic environmental changes, the impact of multiple concurrent stressors, and region adaptation strategies. The models currently in use, such as that developed by Vugrin et al. in 2011 [39] or CIRAF, introduced by Singhal et al., [41] focus primarily on static resilience metrics or specific geographical areas, without incorporating multiple scenarios of infrastructures exposed to similar hazards. In contrast, our model integrates both infrastructural and environmental variables with real-time data inputs, offering enhanced predictive accuracy and dynamic resilience assessment under shifting conditions. Additionally, compared to existing ones, our model demonstrates a scenario applicability over 70% (see case study results), where multiple risk factors, such as landslides and floods, occur concurrently.

3.3. Case Study

Burundi has a road network of nearly 12,000 km. According to the Roads Office, the internal road network consists of 4456 km of classified roads (i.e., maintained by the Roads Office), including 22 national roads, totaling 1952 km, and 91 provincial roads, with a total length of 2522 km. To examine the practical use of the proposed road resilience assessment framework, this study selected the case of National Road 9 in Burundi, which has recently been affected by landslides triggered by heavy rains. These landslides blocked the roadway and halted traffic during the time of damage. In recent years, Burundi has experienced more than 446 natural disasters, including torrential rains, floods, landslides, and violent winds. On the 18^th of March 2023, a disaster involving heavy rain occurred, resulting in significant damages to society and infrastructures, including NR9. The land on which the road is laid and the ground supporting the road slid, causing a section of the roadside to collapse. Figure 8a and b represent the affected part of the road at Mileage points MP15 and MP28, respectively.

3.3.1. National Road 9 Recovery Process

One day after the disaster (on the morning of the 19th of March 2023), a meeting organized by the Agence Routière du Burundi (ARB, a government institutional agency for road infrastructure) was held. The purpose of the meeting was to take swift action to rapidly recover the damaged road area and reconstruct the affected road sections. The meeting featured road staff, consisting of engineers and disaster management experts. Thus, the following key actions were decided by the end of the meeting:

• Visiting the site and identifying the most affected area of road infrastructure;
• Conducting a geotechnical investigation of the affected road area;
• Carrying out an emergency feasibility study regarding the stabilization of threatened points;
• Identifying and quickly mobilizing potential resource needs;
• Following the existing national disaster management and risk reduction plan to successfully restore road traffic in an efficient and speedy manner.

The roadmap of the actions adopted to recover NR9 from the landslide of the 18th of March 2023 and the road damages caused by heavy rains is presented in Figure 9.

3.3.2. Practical Application and Integration of the RIRA Framework with the Success Performance Factors Implemented for Case Road Recovery

By aligning the RIRA framework with performance factors (

Table 6. Performance factors of NR9 recovery.

	Case Study Recovery Performance Factors	Description	Verification
1	Geotechnical investigation of the affected road area	This consists of laboratory tests and an analysis of the soil characteristics, inquiring about road base and surface conditions (properties, moisture content, porosity, permeability, etc.)	RIRA factors RO11, RO12, RO13, RO14, RO21, RO22, RO23, and RO24 were involved in the road case study soil tests
2	Road design	The task of designing the road consists of geometric and topographic alignment, presenting the recovered road layout. Thus, road facilities such as drainage and signal systems are presented in this step.	Consequently, RO31, RO32, RO34, RO35, RO41, RO42, RO43, RO44, RO51, RO52, RO64, and IN33 were designed and recovered.
3	Resource identification and allocation	The Burundi Road Agency (ARB) owns some human (engineers, labor, etc.), financial, equipment (excavation machines, etc.), and material resources. In the case of a shortage of resources, the agency has rental contracts with private agencies working in the construction sector. In addition, Burundi has a laboratory well equipped to carry out the necessary geotechnical studies. During the recovery process of NR9, the necessary resources were available.	Factors RE11, RE12, RE13, and RE14 of the RIRA framework were verified to influence the recovery process of NR9.
4	Recovery processes and management (organization)	The ARB has staff organized into different departments to manage construction projects in an accurate manner. Thus, project monitoring, site inspection, surveillance, and decision making were carried out well during the recovery of NR9.	Factors RE21, RE22, RE23, RE24, RE41, RE42, RE43, RE44, and RE45 influenced NR9 recovery.

) applied during road case study recovery, the recovery team could achieve not just a functional road, but also one that met long-term objectives like resilience, safety, and cost-effectiveness. This systematic application facilitated the project’s overall success, demonstrating the framework’s ability to adapt to dynamic challenges while ensuring stakeholder satisfaction.

3.3.3. Discussion of the Results

The results recorded in Table 5 show that all 13 factors implemented for resource identification/allocation and recovery process/management during the recovery of the national road used as a case study fell within the 17 road recoverability factors presented in the RIRA framework. Therefore, the performance factors in the NR9 case study represented 76% of RIRA framework recoverability factors. Additionally,19 success factors for NR9 recovery were among the 27 robustness factors included in the road infrastructure resilience assessment framework. Consequently, the performance factors utilized for NR9 recovery corresponded to 70% of RIRA framework robustness factors. Considering the impact of the RIRA framework factors on the performance factors implemented for NR9 recovery, it can be concluded that the road infrastructure resilience assessment framework proposed in this study was reliable for NR9 recovery.

However, even though the framework complied with the recovery process in the case study process, during its implementation, we faced several challenges, including technical limitations with legacy systems, data integration issues due to inconsistencies in data quality, and difficulties coordinating between diverse stakeholders. These challenges extended the implementation timeline and required additional resources for problem solving. We also noted that the case study lacked intelligent systems, so the intelligence dimension factors of the RIRA framework could not be applied to the NR9 recovery plan. Figure 10a,b depict the state of NR9 after recovery at MP15 and MP28, respectively.

4. Conclusions and Future Works

To expand curve-based infrastructure resilience assessments, this paper proposed a novel resilience assessment method named “New Infrastructure Resilience Assessment Curve (NIRAC)”. While the existing literature typically provides a single-scenario infrastructure resilience assessment curve, the NIRAC expands on this by utilizing multi-infrastructure scenarios, offering a more comprehensive understanding of infrastructure performance and its environmental implications under similar hazards. Additionally, this paper also presented a comprehensive road resilience assessment framework named the RIRA framework, based on the identified road infrastructure resilience factors. To ensure the efficiency of the factors, a systematic survey was distributed among road reconstruction and disaster management experts to collect data. A statistical analysis on the experts’ perceptions revealed that all the identified factors presented high mean scores and significance levels, so they were suitable for assessing the resilience of road infrastructure.

For a better understanding of how to manage and improve road resistance during disasters, to anticipate the challenges and opportunities at each stage, thus ensuring a more resilient road transportation infrastructure, this paper classified resilience factors according to the phases of a disaster and mapped them onto a road performance curve. To validate the effectiveness of the proposed road infrastructure resilience assessment framework, it was applied to a real-world case study involving National Road number 9 in Burundi, which suffered damage due to flooding caused by heavy rains, highlighting the importance of sustainable recovery efforts. The evaluation results proved that the RIRA framework was suitable for assessing the recovery of the studied road.

However, it is important to acknowledge certain limitations of this research. Both the NIRAC and RIRA frameworks primarily focus on a theoretical approach to assess the resilience and sustainability of critical infrastructures. Additionally, the chosen case study may not fully represent the modern, smart roads found in today’s infrastructural landscape. Furthermore, this study did not quantitatively apply the developed NIRAC in a case study for verification purposes. For a deep understanding of the assessment of critical infrastructural resilience, future research endeavors could expand upon the methods proposed in this paper and apply them to specific, complex, and smart road cases. By incorporating empirical analyses and applying the approach presented in this research paper to multiple, modern case studies, such research could provide deeper insights into sustainable resilience assessment.

As environmental challenges will continue to intensify, this paper recommends the effective implementation and improvement of environmental impact assessments (EIAs) for balancing economic growth sustainability with environmental preservation in future research. This paper further encourages comprehensive stakeholder engagement to ensure that diverse perspectives and concerns are addressed throughout the decision making process by involving all relevant parties, such as local communities, government agencies, industry experts, and environmental groups, as stakeholder engagement fosters transparency, trust, and collaboration. This inclusive approach will not only help identify potential challenges early on but also enhance the social acceptability of projects, leading to more informed, balanced, and equitable outcomes. Ultimately, comprehensive stakeholder engagement strengthens project sustainability and supports long-term community and environmental well-being.

Future works could further expand this research’s foundation by integrating additional dimensions and factors that may arise in different contexts or under evolving conditions. This would enable disaster managers and policymakers to gain insights into the state of infrastructural resilience and identify potential measures for its enhancement.