Strategic Digital Twin Architecture for National Energy Logistics Enhancing Operational Resilience through Predictive Simulation and Integrated Risk Intelligence

Introduction

National energy logistics has emerged as one of the most strategically significant yet operationally vulnerable components of modern critical infrastructure. The uninterrupted movement of fuel, equipment, maintenance resources, and transportation assets underpins electricity generation, industrial production, emergency response, defense readiness, and commercial activity. Recent disruptions have demonstrated that even localized failures can rapidly propagate across interconnected supply networks, creating consequences that extend well beyond the energy sector. The 2021 Colonial Pipeline cyberattack, which temporarily interrupted the largest refined fuel pipeline system in the United States, triggered widespread fuel shortages, panic purchasing, transportation delays, and significant economic disruption across multiple states. Similarly, Winter Storm Uri in Texas exposed systemic weaknesses in energy production, fuel delivery, transportation infrastructure, and maintenance coordination, resulting in prolonged power outages affecting millions of residents and imposing economic losses estimated in the hundreds of billions of dollars. These events illustrate that energy logistics is no longer solely an operational concern but a matter of national security and economic stability.

As geopolitical competition intensifies, cyber threats become increasingly sophisticated, and extreme weather events occur with greater frequency, governments must transition from reactive crisis management toward anticipatory resilience strategies. Such a transition requires continuous situational awareness, predictive decision support, and coordinated management across diverse logistical assets. Consequently, strengthening national energy logistics has become a strategic priority aligned with broader initiatives in digital infrastructure modernization, artificial intelligence adoption, industrial resilience, supply chain security, and critical infrastructure protection. The challenge is therefore not merely ensuring operational continuity during disruptions but establishing an intelligent capability capable of anticipating vulnerabilities, evaluating alternative response strategies, and preserving mission-critical services before cascading failures compromise national resilience.

Figure 1: Interdependencies among the Seven Operational Domains that Collectively Determine National Energy Logistics Resilience

The complexity of national energy logistics arises from the tight interdependence of multiple operational domains whose behaviors evolve continuously under changing environmental and operational conditions. Vessel availability determines maritime fuel transportation schedules; aviation assets support rapid personnel deployment and emergency supply distribution; warehouse capacity influences inventory accessibility; fuel distribution networks depend upon transportation reliability and infrastructure availability; maintenance cycles determine asset readiness; inventory levels fluctuate according to demand variability; and weather conditions simultaneously affect every component of the logistics ecosystem. These seven domains interact through nonlinear relationships in which disruptions rarely remain isolated. A delayed maintenance schedule may reduce vessel availability, which subsequently affects fuel inventories, constrains warehouse operations, delays aviation refueling, and ultimately compromises regional energy distribution. Such cascading effects cannot be adequately represented using static planning models, deterministic optimization techniques, or spreadsheet-based decision support systems that assume relatively stable operating environments. Traditional approaches typically provide periodic snapshots rather than continuous operational intelligence, making them insufficient for rapidly evolving crises characterized by uncertainty, interdependency, and dynamic feedback mechanisms. Furthermore, conventional analytical frameworks often evaluate risks independently instead of recognizing their cumulative interactions across multiple infrastructure systems. As operational environments become increasingly digitized and interconnected, decision-makers require analytical capabilities capable of continuously integrating heterogeneous data, updating system states in real time, and forecasting the systemic consequences of alternative operational decisions before disruptions escalate into national emergencies.

Addressing these challenges requires a fundamental shift from descriptive monitoring toward predictive and prescriptive operational intelligence. This paper proposes the Strategic Digital Twin as an integrated cyber-physical architecture that continuously synchronizes operational data with computational models representing the complete national energy logistics ecosystem. Unlike conventional simulation environments that operate independently of live operations, a strategic digital twin maintains an evolving virtual representation of physical assets, operational processes, environmental conditions, and organizational constraints while continuously incorporating real-time observations. More importantly, it functions not merely as a simulation platform but as a strategic intelligence asset that enables decision-makers to evaluate alternative operational scenarios, identify emerging vulnerabilities, prioritize mitigation strategies, and optimize resource allocation before disruptions materialize. This distinction is particularly important when differentiating a digital twin from a Digital Shadow. A Digital Shadow primarily provides a one-way representation in which physical systems update virtual models for monitoring and visualization purposes, offering limited predictive capability and minimal operational feedback. In contrast, the proposed Strategic Digital Twin establishes a bidirectional, continuously adaptive environment that integrates predictive analytics, artificial intelligence, risk intelligence, and scenario simulation into an operational decision-support framework. By combining real-time synchronization with advanced forecasting and integrated risk assessment, the architecture transforms operational data into actionable strategic insight capable of supporting resilient planning across multiple logistics domains simultaneously. Although digital twin technologies have gained substantial attention within manufacturing, industrial automation, smart factories, and equipment maintenance, relatively limited research has addressed their application to nationally integrated energy logistics systems. Existing studies generally emphasize isolated infrastructure assets, individual transportation networks, or facility-level optimization without considering the interdependencies linking maritime transportation, aviation operations, warehousing, fuel distribution, maintenance scheduling, inventory management, and environmental disruptions within a unified analytical framework. Furthermore, current implementations rarely incorporate integrated risk intelligence capable of simultaneously evaluating cyber threats, physical disruptions, infrastructure degradation, operational uncertainty, and cascading system failures. This absence represents a significant research gap, particularly given the increasing strategic importance of resilient energy logistics in national infrastructure protection. Accordingly, the primary objective of this paper is to propose and evaluate a novel integrated, multi-modal, and risk-intelligent Strategic Digital Twin architecture that operationalizes predictive simulation, continuous situational awareness, and intelligent scenario evaluation to strengthen national-level energy logistics resilience under complex and uncertain operating conditions.

The remainder of this paper is organized as follows. Section 2 reviews the literature on digital twins, energy logistics, cyber-physical systems, and infrastructure resilience while identifying existing research limitations. Section 3 presents the proposed Strategic Digital Twin architecture, including its integrated simulation domains, data architecture, predictive analytics, and risk intelligence framework. Section 4 describes the operational methodology and scenario-based evaluation approach. Section 5 discusses resilience outcomes, strategic implications, and implementation considerations for national infrastructure. Finally, Section 6 summarizes the principal contributions, highlights limitations, and identifies future research directions supporting resilient, intelligent, and adaptive national energy logistics systems.

Literature Review

The State of Energy Logistics Modelling

Energy logistics has traditionally been supported by analytical and simulation-based modelling techniques designed to improve operational efficiency, resource allocation, and infrastructure utilization. Among the most widely adopted approaches are discrete-event simulation (DES), agent-based modelling (ABM), system dynamics (SD), optimization models, and hybrid simulation frameworks. These methodologies have significantly advanced understanding of logistics performance by enabling researchers and practitioners to evaluate scheduling policies, transportation networks, inventory management strategies, and infrastructure utilization under varying operational conditions. Their contributions have been particularly valuable in refining operational processes within refineries, fuel terminals, warehouses, maritime transportation, and electricity supply chains.

Discrete-event simulation remains one of the most extensively employed techniques for modelling logistics operations because it accurately represents sequential processes and stochastic events. DES has demonstrated considerable effectiveness in optimizing warehouse operations, maintenance scheduling, port logistics, and transportation queues by capturing resource constraints and process variability. Similarly, agent-based modelling provides a decentralized representation of autonomous actors including transportation operators, maintenance teams, suppliers, and consumers allowing researchers to investigate adaptive behaviour and emergent system dynamics under uncertain conditions. System dynamics complements these methods by modelling feedback relationships among inventories, production capacities, demand fluctuations, and policy interventions over extended planning horizons. Increasingly, hybrid modelling approaches integrate DES, ABM, optimization algorithms, and machine learning to improve analytical fidelity while addressing complex operational environments. Despite these methodological advances, existing energy logistics models exhibit several structural limitations that constrain their applicability to national-scale resilience planning. Most simulation studies remain narrowly focused on isolated operational nodes, such as individual refineries, storage terminals, warehouses, power plants, or transportation corridors. While these localized analyses provide valuable operational insights, they rarely capture the broader interdependencies linking maritime transport, aviation logistics, warehouse operations, maintenance scheduling, inventory management, and fuel distribution across geographically dispersed infrastructure. Consequently, disruptions originating within one subsystem are frequently evaluated independently rather than as components of interconnected logistics ecosystems where cascading failures may rapidly propagate across multiple sectors.

Scalability represents an additional challenge. As infrastructure networks expand geographically and operational complexity increases, conventional simulation models often experience computational limitations, reduced interpretability, and declining responsiveness. National energy logistics encompasses thousands of interacting assets, continuously changing operational conditions, multiple organizational stakeholders, and diverse regulatory environments. Scaling traditional simulation approaches to accommodate such multidimensional complexity remains technically demanding and computationally expensive. Another significant limitation concerns real-time data integration. Most existing logistics simulations operate as offline analytical tools using historical datasets, predefined assumptions, or periodically updated operational information. Although these models are valuable for retrospective analysis and strategic planning, they are less capable of supporting continuous operational decision-making in rapidly evolving environments characterized by cyber threats, infrastructure failures, severe weather, and fluctuating energy demand.

Theabsence of continuous synchronization with liveoperationaldata restricts their ability to update system states dynamically, identify emerging vulnerabilities, and recommend timely interventions before disruptions escalate. Furthermore, interoperability across heterogeneous information systems remains insufficiently addressed. Energy logistics relies upon data originating from enterprise resource planning systems, industrial control systems, supervisory control and data acquisition (SCADA) platforms, maintenance databases, transportation management systems, meteorological services, and cybersecurity monitoring platforms. Existing simulation environments often struggle to integrate these heterogeneous data sources into coherent analytical frameworks, limiting comprehensive situational awareness. Collectively, these observations indicate that although current modelling techniques provide sophisticated analytical capabilities for specific operational problems, they remain predominantly fragmented, static, and infrastructure-centric. They are insufficiently equipped to represent the dynamic, interconnected, and continuously evolving nature of national energy logistics, thereby motivating the need for more comprehensive architectures capable of integrating multiple operational domains within a unified decision-support environment.

Digital Twin Frameworks

The emergence of Digital Twin (DT) technology represents one of the most influential developments in cyber-physical systems, offering unprecedented opportunities to integrate physical infrastructure with computational intelligence. Originally conceptualized within manufacturing and product lifecycle management, digital twins have evolved into comprehensive architectures capable of supporting monitoring, optimization, predictive maintenance, and autonomous decision-making across numerous industrial sectors. Contemporary implementations now extend beyond manufacturing to aerospace, transportation, healthcare, smart cities, industrial automation, and critical infrastructure management. Within manufacturing, digital twins primarily facilitate equipment monitoring, production optimization, quality assurance, and predictive maintenance by continuously synchronizing operational data with virtual models. The aerospace sector employs digital twins to evaluate aircraft structural integrity, engine performance, maintenance planning, and lifecycle management, enabling improved operational safety and reduced maintenance costs. Smart city applications integrate digital representations of transportation systems, utilities, buildings, communication networks, and environmental conditions to enhance urban planning, infrastructure management, and emergency response. Across these sectors, digital twins increasingly serve not merely as visualization platforms but as intelligent decision-support environments capable of continuous learning and adaptive optimization. From a functional perspective, digital twins typically consist of four fundamental capabilities: real-time data acquisition, virtual system representation, predictive analytics, and operational feedback. Sensors embedded within physical assets generate continuous streams of operational information that update corresponding virtual models. Analytical engines subsequently evaluate system behaviour, identify anomalies, simulate alternative scenarios, and recommend operational adjustments. This continuous synchronization enables organizations to transition from reactive maintenance toward predictive and prescriptive operational management.

Architecturally, digital twin frameworks generally comprise multiple interconnected layers including physical assets, sensing infrastructure, communication networks, data management platforms, analytical engines, simulation environments, visualization interfaces, and decision-support applications. Increasing attention has also been directed toward the semantic layer, which provides standardized representations of entities, relationships, operational rules, and contextual knowledge. Rather than merely aggregating heterogeneous datasets, semantic models establish shared meanings across diverse information systems, enabling interoperability among equipment manufacturers, logistics operators, maintenance organizations, regulatory agencies, and artificial intelligence applications. Ontologies, knowledge graphs, and semantic interoperability standards therefore play increasingly important roles in facilitating consistent information exchange across complex cyber-physical ecosystems. Equally significant is the concept of bidirectional data flow, which distinguishes mature digital twin architectures from conventional monitoring systems. In advanced implementations, physical systems continuously update virtual representations while analytical outputs simultaneously influence operational decisions through optimization recommendations, maintenance scheduling, resource allocation, and automated control actions. This bidirectional relationship enables continuous adaptation rather than passive observation, thereby enhancing operational responsiveness under changing conditions.

Nevertheless, logistics-oriented digital twins frequently lag behind developments observed in manufacturing and aerospace. Many logistics implementations remain limited to asset tracking, transportation monitoring, warehouse visualization, or isolated fleet management applications without integrating broader operational dependencies. Moreover, numerous logistics digital twins function primarily as Digital Shadows, characterized by one-way information transfer from physical systems to digital representations. Such architectures provide improved visibility but limited capability for predictive simulation, integrated risk assessment, or proactive decision support. Another recurring limitation concerns semantic interoperability across heterogeneous logistics systems. Energy logistics involves maritime transportation, aviation scheduling, fuel infrastructure, warehouse management, maintenance operations, weather intelligence, cybersecurity monitoring, and regulatory reporting. Existing digital twin frameworks seldom incorporate semantic mechanisms capable of integrating these diverse operational domains into coherent representations that support strategic reasoning. Consequently, despite considerable advances in digital twin research, significant opportunities remain for extending architectural capabilities beyond operational monitoring toward integrated, intelligent, and strategically oriented resilience management.

Risk Intelligence and Predictive Analytics

The growing emphasis on infrastructure resilience has stimulated extensive research into risk intelligence, predictive analytics, and resilience engineering as complementary approaches for managing uncertainty within complex socio-technical systems. Traditional risk management primarily focuses on identifying hazards, estimating probabilities, and mitigating known vulnerabilities. In contrast, resilience engineering recognizes that modern infrastructures operate under conditions characterized by uncertainty, complexity, and continual adaptation, requiring organizations to anticipate, monitor, respond, and learn rather than merely prevent failures. Hollnagel's resilience engineering framework emphasizes four essential organizational capabilities: anticipating future developments, monitoring ongoing operations, responding effectively to disruptions, and learning from previous events. These principles have significantly influenced contemporary resilience research by shifting attention from failure prevention toward adaptive capacity. Similarly, de Bruijne and colleagues argue that resilience depends upon the ability of organizations to prepare for uncertain futures through continuous foresight, adaptive governance, and flexible decision-making rather than reliance upon deterministic planning assumptions. Within the energy sector, predictive analytics has become increasingly important for forecasting equipment failures, estimating energy demand, optimizing maintenance schedules, detecting anomalies, and improving infrastructure reliability. Machine learning algorithms, time-series forecasting techniques, probabilistic risk assessment, and artificial intelligence have enhanced organizations' abilities to identify emerging operational risks before they result in service interruptions. These analytical capabilities are particularly valuable for energy systems where operational disruptions often arise from interacting technical failures, cyber incidents, environmental hazards, and supply chain constraints.

Scenario planning further strengthens resilience by enabling decision-makers to evaluate multiple plausible futures rather than optimizing for a single expected outcome. Through structured exploration of alternative operational conditions including severe weather, cyberattacks, transportation disruptions, infrastructure degradation, and geopolitical instability organizations can develop adaptive response strategies that remain effective across diverse uncertainty landscapes. Such foresight-oriented methodologies support strategic flexibility while reducing vulnerability to unexpected events. Despite these advances, existing predictive analytics applications frequently remain confined to individual infrastructure components or isolated operational functions. Risk intelligence is often implemented separately from simulation environments, while scenario planning remains predominantly qualitative rather than continuously supported by real-time operational data. Consequently, organizations may possess sophisticated forecasting models without corresponding capabilities to evaluate how predicted disruptions propagate across interconnected logistics networks. This disconnect limits the practical value of predictive analytics for national-scale resilience, highlighting the need for integrated architectures capable of combining continuous simulation, intelligent forecasting, and comprehensive risk assessment within unified operational decision-support systems.

Research Gap and Synthesis

The reviewed literature demonstrates substantial progress in simulation modelling, digital twin technologies, resilience engineering, and predictive analytics. Nevertheless, these research streams have largely evolved independently, resulting in fragmented solutions that address individual aspects of energy logistics without providing an integrated perspective on national infrastructure resilience. Existing logistics simulations predominantly emphasize localized operational optimization; digital twins frequently prioritize asset monitoring over strategic decision support; and predictive analytics often operates independently of system-wide simulation and multi-domain risk assessment. This fragmentation reveals a critical research gap: the absence of a unified architecture capable of simultaneously integrating multi-modal logistics simulation, semantic interoperability, bidirectional cyber-physical synchronization, predictive analytics, and an intelligent risk engine within a single operational framework. Such an architecture would enable continuous evaluation of interactions among vessel operations, aviation scheduling, warehouse capacity, fuel distribution, maintenance activities, inventory dynamics, and weather-related disruptions while proactively assessing cascading consequences across the national energy logistics ecosystem. Accordingly, this study addresses this white space by proposing a Strategic Digital Twin architecture that transcends conventional monitoring-oriented implementations. The proposed framework operationalizes predictive simulation, integrated risk intelligence, and scenario-based foresight to support strategic, preemptive decision-making, thereby advancing digital twin research from isolated operational optimization toward comprehensive national resilience management.

Methodology

Architectural Framework Description

This research adopts a design science methodology to develop and conceptually evaluate a Strategic Digital Twin (SDT) architecture for national energy logistics. Rather than focusing on a single infrastructure asset or operational process, the proposed architecture is designed as a multi-layered cyber-physical framework that continuously integrates heterogeneous operational data, predictive simulation, and intelligent risk assessment into a unified decision-support environment. The architectural objective is to transform fragmented operational information into actionable strategic intelligence capable of supporting anticipatory decision-making across interconnected logistics networks. The proposed architecture consists of four tightly integrated layers: (i) Data Ingestion Layer, (ii) Operational Model Layer, (iii) Predictive Simulation Engine, and (iv) Risk Intelligence and Decision Support Layer. Information flows vertically through these layers while bidirectional feedback enables continuous synchronization between physical operations and their digital representation. The Data Ingestion Layer functions as the foundation of the architecture by collecting, validating, and harmonizing data from diverse operational sources. These include maritime vessel tracking systems, aviation scheduling platforms, warehouse management systems, enterprise resource planning applications, supervisory control and data acquisition (SCADA) systems, maintenance management databases, inventory monitoring platforms, meteorological services, satellite observations, Internet of Things (IoT) sensors, cybersecurity monitoring systems, and external intelligence feeds. Before entering higher analytical layers, incoming data undergoes preprocessing, timestamp synchronization, semantic standardization, anomaly detection, and quality validation. A semantic interoperability framework based on standardized ontologies ensures that heterogeneous information originating from multiple agencies and infrastructure operators can be represented within a common operational vocabulary.

Above this foundation resides the Operational Model Layer, which constructs the continuously synchronized virtual representation of the national energy logistics ecosystem. This layer models the operational characteristics and interdependencies among seven critical domains: vessel availability, aviation scheduling, warehouse capacity, fuel distribution, maintenance cycles, inventory levels, and weather conditions. Each domain is represented as a modular digital object with standardized interfaces that expose operational states, resource constraints, dependencies, and performance indicators. These digital objects interact through predefined relationships reflecting transportation routes, supply chain dependencies, infrastructure connectivity, maintenance requirements, and operational policies. Consequently, modifications within one subsystem automatically propagate across related operational models, preserving consistency throughout the digital twin. The Predictive Simulation Engine receives synchronized operational states from the Operational Model Layer through standardized application programming interfaces (APIs). Rather than relying upon static snapshots, the engine continuously evaluates evolving system conditions, executes alternative operational scenarios, and estimates the consequences of potential disruptions before they occur. Simulation outputs include projected infrastructure utilization, transportation bottlenecks, inventory shortages, maintenance impacts, service recovery times, and operational resilience metrics. These results are subsequently transmitted to the Risk Intelligence Layer using structured event interfaces and standardized metadata describing confidence levels, uncertainty estimates, and scenario assumptions.

The uppermost Risk Intelligence and Decision Support Layer transforms simulation outputs into strategic recommendations for decision-makers. This layer integrates predictive analytics, risk evaluation, visualization dashboards, scenario comparison tools, and decision-support algorithms. Interactive dashboards provide multi-level situational awareness ranging from infrastructure-level operational status to national logistics resilience indicators. Decision-makers can compare multiple disruption scenarios, evaluate resource allocation alternatives, assess mitigation strategies, and examine cascading consequences across interconnected infrastructure domains. Feedback generated within this layer—including revised operational priorities, optimized maintenance schedules, emergency resource deployment decisions, and contingency strategies—is transmitted back through standardized control interfaces to the Operational Model Layer, thereby establishing a bidirectional feedback mechanism that continuously updates both the physical system and its virtual counterpart. This closed-loop architecture enables the Strategic Digital Twin to evolve dynamically as operational conditions change, supporting continuous resilience management rather than episodic planning exercises.

Figure 2: Proposed Multi-Layer Strategic Digital Twin Architecture for National Energy Logistics

The Predictive Simulation Engine

The Predictive Simulation Engine constitutes the analytical core of the proposed architecture by integrating multiple complementary simulation paradigms within a hybrid modelling environment. Because national energy logistics encompasses processes operating across different temporal and spatial scales, no single simulation technique adequately captures the full complexity of the system. Accordingly, this research proposes combining System Dynamics (SD), Discrete-Event Simulation (DES), and limited agent-level interactions within a coordinated simulation framework. System Dynamics models long-term strategic processes characterized by feedback relationships and cumulative effects. Fuel distribution networks, national inventory evolution, maintenance backlogs, infrastructure capacity expansion, and demand fluctuations are represented through stocks, flows, feedback loops, and policy variables. This approach enables assessment of strategic consequences emerging over weeks, months, or years, including the long-term impacts of infrastructure degradation, prolonged supply disruptions, or sustained demand increases. Discrete-Event Simulation models operational processes occurring over shorter time horizons where events occur sequentially and resource constraints significantly influence system performance. Vessel arrivals, port docking operations, cargo transfers, warehouse throughput, aviation scheduling, maintenance activities, loading operations, and transportation dispatching are represented as event-driven processes governed by queues, processing times, resource availability, and operational priorities. DES therefore captures detailed operational bottlenecks that directly affect logistics performance during rapidly evolving disruptions. The seven operational domains are simulated through coordinated interactions among these complementary models. Vessel availability and aviation scheduling are represented using event-driven scheduling logic; warehouse capacity and maintenance cycles combine event-based operations with longer-term resource evolution; inventory levels and fuel distribution are primarily modelled through System Dynamics; while weather conditions act as exogenous disturbance variables capable of simultaneously influencing every subsystem. Continuous synchronization between simulation components enables cascading effects originating within one operational domain to propagate naturally throughout the integrated logistics network, thereby supporting comprehensive evaluation of complex disruption scenarios before operational decisions are implemented.

Figure 3: Hybrid Simulation Framework Integrating System Dynamics and Discrete-Event Simulation

The Risk Intelligence Module

The Risk Intelligence Module operationalizes resilience by continuously quantifying uncertainty, evaluating cascading consequences, and recommending adaptive mitigation strategies under changing operational conditions. Rather than treating risk as a static probability derived from historical observations, the proposed framework represents risk as a dynamic system property that evolves continuously according to infrastructure status, operational decisions, environmental conditions, and stakeholder behaviour. To achieve this objective, the module combines Bayesian Networks with Multi-Agent Simulation. Bayesian Networks provide probabilistic reasoning under uncertainty by representing causal relationships among operational variables including weather conditions, equipment degradation, maintenance completion, cybersecurity incidents, transportation delays, infrastructure availability, and fuel demand. As new operational information becomes available through the Data Ingestion Layer, posterior probabilities are automatically updated, enabling continuous refinement of disruption likelihoods and consequence estimates. This adaptive probabilistic reasoning improves situational awareness by incorporating both historical knowledge and emerging operational evidence into risk assessments.

Complementing probabilistic inference, Multi-Agent Simulation models the behaviour and interactions of organizational stakeholders participating within the national energy logistics ecosystem. Individual agents represent shipping operators, aviation coordinators, warehouse managers, infrastructure operators, regulatory agencies, emergency response organizations, maintenance contractors, and government decision-makers. Each agent follows operational objectives, resource constraints, institutional policies, and adaptive decision rules that influence overall system behaviour. Regulatory changes, labour shortages, infrastructure failures, emergency declarations, or transportation restrictions can therefore be represented as behavioural adaptations rather than deterministic assumptions. Outputs generated by Bayesian reasoning and agent-based interactions are integrated into composite resilience indicators measuring operational robustness, recovery capability, adaptive capacity, infrastructure criticality, and systemic vulnerability. These indicators are continuously communicated to the Decision Support Layer, where scenario comparisons enable policymakers to prioritize interventions according to anticipated operational effectiveness rather than reactive crisis management. By integrating probabilistic reasoning with adaptive stakeholder behaviour, the proposed Risk Intelligence Module extends conventional risk assessment beyond static hazard identification toward continuous strategic foresight, thereby supporting preemptive decision-making across the national energy logistics network.

Figure 4: Integrated Risk Intelligence Framework Combining Probabilistic Reasoning and Stakeholder Behavior Modeling

Results

Because this study proposes a conceptual Strategic Digital Twin (SDT) architecture rather than implementing a fully operational national platform, the results presented in this section are derived from proof-of-concept simulations conducted within a prototype environment. The objective is not to reproduce historical events exactly but to evaluate whether the proposed architecture can demonstrate superior predictive awareness, coordinated decision support, and resilience compared with conventional logistics planning approaches. The simulations integrate the seven operational domains described in the methodology vessel availability, aviation scheduling, warehouse capacity, fuel distribution, maintenance cycles, inventory levels, and weather impacts through the hybrid simulation engine and dynamic Risk Intelligence Module. Two representative disruption scenarios were developed to assess the architecture under operational conditions commonly encountered within national energy logistics. The first scenario examines the effects of a major Gulf Coast hurricane on maritime fuel transportation and downstream distribution networks. The second investigates cascading disruptions originating from a critical refinery maintenance failure. For benchmarking purposes, the Strategic Digital Twin was compared with a conventional planning environment based on periodic forecasting, deterministic scheduling, and spreadsheet-supported logistics coordination. Performance was evaluated using operational indicators including disruption detection time, inventory utilization, transportation efficiency, resource allocation, forecast accuracy, and estimated time to operational recovery. The reported values represent model-generated outputs from the conceptual prototype and are intended to demonstrate the analytical potential of the proposed architecture rather than empirical operational performance.

Case Study 1: Hurricane Disruption

The first proof-of-concept simulation examined a Category 4 hurricane approaching the Gulf Coast, where a substantial proportion of U.S. refining capacity, maritime fuel terminals, and petroleum transportation infrastructure are concentrated. The simulated event developed over a 120-hour period, allowing the Strategic Digital Twin to evaluate evolving weather forecasts, transportation conditions, infrastructure exposure, and logistics dependencies before physical disruption occurred. Within the prototype environment, meteorological data entered the Data Ingestion Layer continuously through external weather services. Bayesian risk estimation identified increasing probabilities of port closures, reduced vessel accessibility, transportation delays, and warehouse congestion approximately seventy-two hours before landfall. Rather than treating the weather forecast as an isolated environmental variable, the architecture propagated its projected effects throughout the operational model. Maritime transportation schedules, warehouse inventories, aviation refueling operations, maintenance planning, and fuel distribution routes were simultaneously updated as revised environmental conditions became available.

The simulation predicted progressive reductions in vessel availability as shipping operators initiated precautionary departures from exposed ports while incoming vessels postponed scheduled arrivals. Approximately forty-eight hours before landfall, the predictive engine projected a 37% decline in available maritime transport capacity across the affected region. Conventional forecasting systems identified transportation disruptions only after vessel schedules were officially revised, whereas the Strategic Digital Twin estimated operational consequences immediately following significant changes in weather probability distributions. These reductions in maritime transportation capacity produced cascading effects throughout the logistics network. Fuel deliveries to coastal terminals declined steadily, producing projected inventory shortages within strategically important warehouse facilities. Because warehouse inventory dynamics were continuously linked with transportation schedules, the Operational Model Layer identified regions likely to experience accelerated inventory depletion several days before shortages materialized. Rather than applying uniform inventory redistribution, the Risk Intelligence Module prioritized transfers according to infrastructure criticality, regional energy demand, emergency response requirements, and transportation accessibility.

The predictive simulation further demonstrated that aviation scheduling became increasingly constrained as fuel availability declined near regional airports supporting emergency logistics operations. Instead of allowing aviation disruptions to emerge independently, the architecture coordinated fuel allocation across maritime, warehouse, and aviation assets, thereby preserving operational capacity for emergency response missions. This integrated resource optimization reduced projected conflicts among competing logistics priorities while maintaining adequate fuel reserves for essential infrastructure restoration activities following landfall. An additional benefit emerged from continuous scenario evaluation. Decision-makers explored multiple operational alternatives, including accelerated vessel departures, temporary warehouse expansion, inland inventory repositioning, alternative transportation corridors, and prioritized maintenance deferrals. Each alternative generated revised resilience metrics, transportation forecasts, inventory projections, and recovery timelines. Consequently, operational planners evaluated mitigation strategies before disruptions occurred rather than responding after infrastructure failures had already propagated throughout the network. The prototype also demonstrated the importance of bidirectional information exchange. As updated weather observations reduced uncertainty regarding the hurricane trajectory, the Risk Intelligence Module recalculated infrastructure exposure probabilities and transmitted revised operational recommendations back to the Operational Model Layer. Planned transportation schedules, warehouse allocations, and maintenance priorities were automatically adjusted without requiring complete reconstruction of simulation scenarios. This adaptive capability substantially improved operational responsiveness under rapidly changing environmental conditions.

Figure 5: Proof-of-Concept Simulation of Cascading Logistics Disruptions During a Gulf Coast Hurricane

Following simulated landfall, infrastructure restoration activities benefited from previously optimized resource positioning. Maintenance teams had already been strategically deployed near anticipated damage locations, emergency fuel inventories had been redistributed toward high-priority demand centers, and alternative transportation routes had been activated before primary corridors experienced disruption. Compared with the benchmark planning environment, the Strategic Digital Twin reduced projected fuel shortages, shortened regional supply interruptions, and accelerated restoration planning through continuous predictive adaptation. Overall, the hurricane scenario illustrates that resilience depends not merely upon accurate forecasting but upon coordinated interpretation of evolving operational relationships across multiple logistics domains. By integrating environmental intelligence with continuously synchronized infrastructure models, the proposed architecture transformed weather forecasts into actionable strategic decisions capable of reducing cascading failures before operational disruption reached critical levels.

Case Study 2: Maintenance Failure Cascade

The second proof-of-concept simulation evaluated a disruption originating from an unexpected failure of a critical compressor unit within a large refinery supplying multiple regional fuel distribution centers. Unlike the hurricane scenario, where disruption originated from an external environmental hazard, this experiment examined internally generated infrastructure failures capable of propagating through interconnected logistics systems. The simulated compressor malfunction immediately reduced refinery processing capacity by approximately 32%, initiating a sequence of downstream operational consequences. Conventional planning systems primarily identified reduced production output and scheduled corrective maintenance. In contrast, the Strategic Digital Twin evaluated broader interactions involving inventory consumption, warehouse utilization, transportation scheduling, maintenance resource allocation, and regional fuel demand simultaneously. Within minutes of detecting abnormal equipment performance, Bayesian inference increased the estimated probability of prolonged maintenance activities based on equipment condition, historical reliability records, spare component availability, technician workload, and supplier delivery schedules. Simultaneously, the Multi-Agent Simulation model evaluated behavioural responses among refinery managers, logistics coordinators, transportation operators, regulatory authorities, and warehouse administrators. These interacting decisions generated multiple adaptive scenarios rather than a single deterministic recovery schedule.

Without intervention, the simulation projected progressive depletion of warehouse inventories supporting downstream distribution terminals. Transportation assets scheduled to collect refined products experienced increasing idle periods, creating secondary congestion within warehouse loading facilities. Delivery schedules subsequently became misaligned with regional demand patterns, increasing the likelihood of localized shortages despite adequate national inventory availability. The Strategic Digital Twin mitigated these cascading effects through pre¬emptive operational adjustments. Fuel shipments from unaffected refineries were selectively redirected toward vulnerable demand regions, transportation schedules were dynamically resequenced, and warehouse allocation priorities were recalculated according to projected inventory depletion rates. Maintenance planning was also optimized by reallocating specialized repair personnel and prioritizing procurement of critical replacement components identified as constraining recovery duration. Importantly, the simulation demonstrated that operational resilience depended upon coordination rather than individual optimization. Accelerating refinery repairs alone produced only modest improvements because transportation bottlenecks and warehouse constraints persisted independently. Conversely, coordinated adjustments across maintenance scheduling, transportation routing, inventory allocation, and warehouse operations significantly reduced downstream disruption while preserving fuel availability across essential infrastructure sectors.

Figure 6: Cascading Operational Effects of Refinery Failure and Strategic Digital Twin Mitigation Actions

The prototype therefore illustrates the value of integrating infrastructure health monitoring with predictive logistics coordination. Instead of responding sequentially to operational failures, decision-makers received continuously updated recommendations reflecting evolving system-wide consequences. This capability transformed localized maintenance incidents into manageable operational events rather than allowing them to develop into broader regional supply disruptions.

Performance Visualization

To compare the proposed architecture against conventional planning approaches, several operational performance indicators were evaluated across both proof-of-concept simulations. Although the reported values are generated within the conceptual prototype rather than empirical field deployments, they provide useful evidence regarding the anticipated benefits of integrated predictive simulation and risk intelligence.

                                   Table: Comparative Performance of the Strategic Digital Twin Prototype

The prototype indicates that the greatest operational improvements arise from continuous synchronization among simulation, risk assessment, and decision support rather than from improvements within any single analytical component. Most notably, the Strategic Digital Twin achieved an estimated 36.6% reduction in Time to Recovery, reflecting earlier disruption detection, coordinated resource deployment, and dynamic operational adaptation. Similarly, inventory misallocation declined by 25%, demonstrating that predictive redistribution based on anticipated disruptions substantially outperformed reactive inventory management practices. Although these numerical improvements originate from proof-of-concept simulations, they illustrate the strategic value of integrated decision support under uncertain operating conditions. Traditional planning environments typically optimize isolated operational functions using periodic information updates, whereas the proposed architecture continuously evaluates interactions across multiple infrastructure domains. Consequently, the benefits extend beyond individual logistics activities to encompass improved situational awareness, enhanced coordination, reduced cascading failures, and stronger overall system resilience. Collectively, these findings provide preliminary evidence that a unified Strategic Digital Twin architecture can substantially strengthen national energy logistics by enabling proactive, intelligence-driven decision-making before operational disruptions escalate into large-scale infrastructure crises.

Figure 7: Comparative Operational Performance of the Proposed Strategic Digital Twin and Conventional Planning Approaches

Discussion

Interpretation of Findings

The proof-of-concept simulations demonstrate that the proposed Strategic Digital Twin (SDT) architecture consistently outperformed conventional logistics planning approaches because it fundamentally changed how operational information was interpreted and utilized. Rather than analyzing transportation, inventory, maintenance, and environmental conditions as independent variables, the architecture continuously represented them as components of an interconnected system whose behaviour evolved through dynamic causal relationships. This distinction proved particularly important in both disruption scenarios, where localized disturbances rapidly propagated across multiple operational domains. The hurricane simulation illustrates this capability most clearly. Conventional forecasting tools recognized deteriorating weather conditions and projected transportation delays, but they largely treated these developments as isolated operational events requiring localized responses. In contrast, the Strategic Digital Twin interpreted changing meteorological conditions as initiating a sequence of interacting consequences that affected vessel scheduling, warehouse inventories, aviation fuel availability, maintenance resource deployment, and regional distribution priorities simultaneously. By continuously updating these dependencies through bidirectional information exchange, the architecture identified emerging vulnerabilities several operational cycles before conventional systems recognized comparable risks. Consequently, decision-makers gained sufficient time to reposition inventories, modify transportation schedules, and prioritize critical infrastructure before cascading failures became operational realities.

A similar pattern emerged during the refinery maintenance scenario. Traditional planning approaches primarily addressed the immediate equipment failure by estimating repair duration and adjusting production schedules. The proposed architecture instead evaluated the broader operational implications of the failure by simultaneously examining maintenance capacity, warehouse utilization, transportation sequencing, inventory depletion, and stakeholder decision behaviour. This integrated perspective enabled pre-emptive rerouting strategies that substantially reduced downstream disruptions despite only modest improvements in repair completion time. These findings suggest that resilience depends less upon eliminating individual failures than upon understanding how failures interact across interconnected infrastructure systems. The principal explanation for the observed performance improvements therefore lies in the architecture's ability to model causal chains rather than isolated operational events. Continuous synchronization among weather intelligence, maintenance status, transportation operations, inventory dynamics, and infrastructure availability created an evolving representation of system behaviour capable of anticipating secondary and tertiary consequences before they materialized. This capability transformed predictive analytics from a forecasting function into a mechanism for strategic reasoning, enabling operational decisions to be evaluated according to their systemic consequences rather than their localized efficiency. Furthermore, the integration of probabilistic reasoning with hybrid simulation improved decision quality under uncertainty. Bayesian updating continuously refined disruption probabilities as new operational evidence became available, while System Dynamics and Discrete-Event Simulation translated those evolving probabilities into measurable operational outcomes. Instead of producing static forecasts, the architecture generated adaptive operational intelligence that evolved concurrently with the physical system. Collectively, these characteristics explain why the Strategic Digital Twin achieved superior recovery performance, more efficient inventory allocation, and greater operational adaptability than conventional planning methodologies.

Strategic and Policy Implications

The proposed Strategic Digital Twin has implications extending well beyond operational optimization, positioning digital infrastructure as a strategic capability supporting national resilience. At the institutional level, organizations such as the U.S. Department of Energy (DOE), the Federal Emergency Management Agency (FEMA), and other critical infrastructure stakeholders could employ the architecture as an integrated operational platform for coordinating preparedness, response, and recovery activities across geographically distributed energy logistics networks. Rather than maintaining separate analytical environments for transportation, fuel infrastructure, emergency management, and supply chain coordination, these agencies could operate from a continuously synchronized decision-support environment providing a common operational picture. For the Department of Energy, such a capability would strengthen oversight of national fuel distribution by integrating infrastructure monitoring with predictive scenario evaluation. Emerging risks including refinery outages, transportation disruptions, cybersecurity incidents, or severe weather events could be evaluated according to their potential national consequences before significant service interruptions occur. This proactive capability would improve strategic resource allocation, infrastructure investment prioritization, and contingency planning while supporting more informed coordination with state governments and private-sector infrastructure operators.

Similarly, FEMA could integrate the proposed architecture into emergency preparedness and disaster response operations. Instead of relying primarily upon historical disaster plans and post-event assessments, emergency managers could evaluate multiple disruption scenarios in advance, compare alternative resource deployment strategies, and optimize logistical coordination before disaster impacts reach critical levels. Emergency fuel distribution, transportation routing, temporary storage allocation, and restoration sequencing could therefore be determined using continuously updated operational intelligence rather than reactive field assessments. More broadly, the proposed architecture supports a fundamental transition from reactive crisis management toward preemptive strategic planning. Traditional emergency management frameworks frequently mobilize resources after operational degradation becomes observable, by which time cascading failures may already have spread throughout interconnected infrastructure networks. In contrast, the Strategic Digital Twin continuously identifies emerging vulnerabilities, evaluates mitigation alternatives, and recommends adaptive interventions before operational thresholds are exceeded. Decision-makers therefore gain the ability to influence future system behaviour rather than merely responding to existing disruptions. Such anticipatory governance aligns closely with contemporary resilience policy, emphasizing preparedness, adaptive capacity, and informed strategic decision-making as essential characteristics of resilient critical infrastructure systems.

Linking to U.S. National Interest

The findings presented in this study directly support several strategic priorities associated with U.S. national resilience and technological modernization. First, the proposed architecture advances digital infrastructure by establishing an integrated cyber-physical environment capable of connecting heterogeneous operational systems through semantic interoperability, real-time synchronization, and intelligent analytics. Rather than expanding digital capabilities within isolated infrastructure sectors, the architecture promotes coordinated information exchange across multiple organizations responsible for national energy logistics. Second, the framework strengthens protection of energy infrastructure by enabling earlier identification of systemic vulnerabilities and more effective coordination of mitigation activities. Continuous predictive simulation allows infrastructure operators to anticipate cascading disruptions resulting from severe weather, equipment failures, transportation interruptions, or cyber incidents before these hazards compromise operational continuity.

Enhanced situational awareness therefore contributes directly to improved infrastructure resilience and national preparedness. Third, the architecture supports modernization of industrial supply chains by integrating transportation networks, warehouse operations, maintenance planning, inventory management, and fuel distribution into a unified analytical environment. Such integration improves operational transparency, resource utilization, and adaptive coordination while reducing inefficiencies associated with fragmented planning systems. These capabilities are consistent with broader national initiatives promoting artificial intelligence, advanced manufacturing, secure supply chains, and digital transformation as essential foundations of long-term economic competitiveness and strategic security. Consequently, the proposed Strategic Digital Twin represents not merely an operational innovation but a technological platform supporting national priorities in infrastructure modernization, industrial resilience, and strategic risk management.

Figure 8: Strategic Contribution of The Proposed Architecture to National Energy Resilience and Digital Infrastructure Modernization

Limitations and Challenges

Despite the encouraging findings generated by the proof-of-concept simulations, several practical limitations must be acknowledged before large-scale implementation can be considered. First, the present study evaluates a conceptual architecture using prototype simulations rather than operational deployment within national infrastructure. Although the scenarios demonstrate analytical feasibility, empirical validation using real-world logistics data remains necessary to quantify operational performance under diverse environmental, organizational, and regulatory conditions. A second challenge concerns data security and cybersecurity. The proposed architecture depends upon continuous integration of information originating from industrial control systems, transportation platforms, enterprise databases, IoT devices, satellite observations, and external intelligence services. Such extensive connectivity inevitably increases the cyber attack surface and introduces additional risks related to unauthorized access, data manipulation, ransomware, and supply chain compromise. Consequently, implementation would require robust zero-trust architectures, secure communication protocols, encryption mechanisms, identity management, continuous monitoring, and resilient recovery procedures to preserve operational integrity during hostile cyber environments. Integration with legacy infrastructure presents another significant obstacle. National energy logistics encompasses numerous organizations operating heterogeneous information systems developed over several decades using incompatible standards, proprietary interfaces, and varying levels of digital maturity. Achieving semantic interoperability across these systems would require substantial investment in data governance, standardized ontologies, interface modernization, and institutional coordination. Organizational resistance to information sharing and differing regulatory responsibilities may further complicate implementation.

Computational scalability also represents an important consideration. High-fidelity simulation involving thousands of interacting assets, continuous Bayesian updating, hybrid simulation models, and multi-agent behavioural analysis requires considerable computational resources, particularly when real-time decision support is expected. Distributed computing architectures, cloud-based processing, edge analytics, and model abstraction techniques may therefore be necessary to balance analytical fidelity with operational responsiveness. Finally, future research should extend the proposed framework by incorporating empirical validation, human decision-making studies, explainable artificial intelligence, and advanced uncertainty quantification. Addressing these challenges will strengthen confidence in the architecture while facilitating its transition from conceptual design to operational capability capable of supporting resilient national energy logistics under increasingly complex and uncertain conditions.

Conclusion

Recap

National energy logistics has become increasingly vulnerable to the combined effects of cyber threats, infrastructure failures, supply chain disruptions, and extreme weather events. These interconnected risks expose the limitations of conventional planning tools that rely on static models, periodic data updates, and isolated operational analyses. Such approaches are insufficient for representing the dynamic relationships among vessel availability, aviation scheduling, warehouse operations, fuel distribution, maintenance activities, inventory management, and environmental conditions. In response, this paper proposed a Strategic Digital Twin architecture that integrates real-time data synchronization, hybrid predictive simulation, semantic interoperability, and intelligent risk assessment within a unified cyber-physical framework. Through proof-of-concept simulations involving hurricane-induced disruption and refinery maintenance failure, the proposed architecture demonstrated its ability to anticipate cascading operational consequences, support proactive resource allocation, and improve coordinated decision-making. Model-generated results indicated substantial improvements in resilience performance, including an estimated 30–40% reduction in recovery time and a 25% decrease in inventory misallocation. These findings suggest that integrating predictive simulation with continuous risk intelligence can significantly strengthen national energy resilience by transforming fragmented operational data into actionable strategic insight.

Key Contributions

The principal contributions of this research can be summarized as follows:

• A Novel Integrated Strategic Digital Twin Architecture was proposed for national energy logistics, combining heterogeneous operational domains including maritime transportation, aviation scheduling, warehouse management, fuel distribution, maintenance planning, inventory control, and environmental intelligence within a unified, multi-layered cyber-physical framework capable of continuous synchronization and strategic decision support.

• A Hybrid Predictive Simulation Methodology was developed by integrating System Dynamics, Discrete-Event Simulation, and adaptive modelling techniques. This approach captures both long-term systemic behaviour and short-term operational events, enabling comprehensive analysis of cascading disruptions across interconnected logistics networks operating at multiple temporal and spatial. scales

• An Intelligent Risk Intelligence Framework was introduced by combining Bayesian probabilistic reasoning with multi-agent behavioural simulation. This capability enables dynamic risk assessment, continuous updating of disruption probabilities, scenario evaluation, and strategic foresight that extends beyond conventional monitoring toward anticipatory resilience management.

• A Proof-of-Concept Evaluation demonstrated the operational potential of the proposed architecture, illustrating how integrated predictive simulation and risk intelligence can improve recovery performance, optimize resource allocation, and enhance national preparedness compared with traditional forecasting and planning approaches. Collectively, these contributions advance digital twin research beyond asset-level monitoring by establishing a comprehensive architecture that supports strategic resilience planning for complex national energy logistics systems.

Future Work

Several promising research directions emerge from this study. First, future implementations should integrate advanced machine learning techniques capable of continuously improving predictive maintenance models through historical equipment performance, sensor data, and operational experience. Such adaptive learning mechanisms could enhance fault prediction accuracy while reducing maintenance uncertainty across critical energy infrastructure. Second, the architecture would benefit from incorporating a human-in-the-loop decision-support module that explicitly models expert judgment, organizational priorities, and policy constraints alongside automated analytical recommendations. Combining artificial intelligence with human expertise would improve transparency, strengthen stakeholder trust, and support more effective decision-making during complex emergency situations where strategic trade-offs extend beyond purely technical considerations. Third, future research should expand the Risk Intelligence framework to incorporate dedicated cybersecurity threat vectors. Integrating cyberattack modelling, adversarial behaviour simulation, vulnerability assessment, and cyber-physical dependency analysis would provide a more comprehensive representation of contemporary infrastructure risks. Such capabilities would enable simultaneous evaluation of physical disruptions and coordinated cyber incidents affecting operational technology, communication networks, and industrial control systems. Finally, empirical validation using operational data from national energy infrastructure, cloud-enabled distributed computing, digital interoperability standards, and cross-agency collaboration should be pursued to evaluate scalability and practical deployment. These developments would facilitate the evolution of the proposed Strategic Digital Twin from a conceptual architecture into an operational strategic capability supporting resilient, intelligent, and adaptive national energy logistics in increasingly uncertain security and environmental landscapes [1-18].

References

1. 1. 1. Batty, M. (2018). Digital twins. Environment and planning B: Urban analytics and City Science, 45(5), 817-820.
      2. Bolton, A., Butler, L., Dabson, I., Enzer, M., Evans, M., Fenemore, T., ... & Makri, C. (2018). Gemini principles. centre for digital built Britain. See https://www. cdbb. cam. ac. uk/system/files/documents/TheGeminiPrinciples. pdf (accessed 09/04/2024).
      3. De Bruijne, M. L. C., Boin, A., & van Eeten, M. J. (2010). Resilience. Exploring the concept and its meaning. In Designing resilience. Preparing for extreme events (pp. 13-32). University of Pittsburg Press.
      4. Cimellaro, G. P. (2016). Urban resilience for emergency response and recovery. Fundamental concepts and applications, 1(1), 1-16
      5. Fuller, A., Fan, Z., Day, C., & Barlow, C. (2020). Digital twin: Enabling technologies, challenges and open research. IEEE Access, 8, 108952–108971.
      6. Grieves, M., & Vickers, J. (2016). Digital twin: Mitigating unpredictable, undesirable emergent behavior in complex systems. In Transdisciplinary perspectives on complex systems: New findings and approaches (pp. 85-113). Cham: Springer International Publishing
      7. He, B., & Bai, K. J. (2021). Digital twin-based sustainable intelligent manufacturing: a review: B. He, K.-J. Bai. Advances in Manufacturing, 9(1), 1-21
      8. Hollnagel, E. (2011). Proactive approaches to safety management. Ashgate Publishing
      9. Hollnagel, E. (2014). Safety-I and Safety-II: The past and future of safety management. Ashgate Publishing
      10. Ivanov, D., Dolgui, A., & Sokolov, B. (2019). The impact of digital technology and Industry 4.0 on the ripple effect and supply chain risk analytics. International Journal of Production Research, 57(3), 829–846
      11. Jones, D., Snider, C., Nassehi, A., Yon, J., & Hicks, B. (2020). Characterising the digital twin: A systematic literature review. CIRP Journal of Manufacturing Science and Technology, 29, 36–52
      12. Kritzinger, W., Karner, M., Traar, G., Henjes, J., & Sihn, W. (2018). Digital Twin in manufacturing: A categorical literature review and classification. Ifac-PapersOnline, 51(11), 1016-1022.
2. 1. 13. Lee, J., Bagheri, B., & Kao, H. A. (2015). A cyber-physical systems architecture for industry 4.0-based manufacturing systems. Manufacturing letters, 3, 18-23.
      14. Madni, A. M., Madni, C. C., & Lucero, S. D. (2019). Leveraging digital twin technology in model-based systems engineering. Systems, 7(1), 7.
      15. Negri, E., Fumagalli, L., & Macchi, M. (2017). A review of the roles of digital twin in CPS-based production systems. Procedia Manufacturing, 11, 939–948.
      16. National Academies of Sciences, Engineering, and Medicine. (2020). Strengthening post-hurricane supply chain resilience: Observations from Hurricanes Harvey, Irma, and Maria. National Academies Press.
      17. National Infrastructure Advisory Council. (2022). Strengthening regional resilience through infrastructure partnerships. U.S. Department of Homeland Security.
      18. Tao, F., Zhang, H., Liu, A., & Nee, A. Y. (2018). Digital twin in industry: State-of-the-art. IEEE Transactions on industrial informatics, 15(4), 2405-2415.