A Theoretical Framework for Graph-based Digital Twins for Supply Chain Management and Optimization

Azmine Toushik Wasi, Mahfuz Ahmed Anik, Abdur Rahman, Md. Iqramul Hoque, MD Shafikul Islam, Md Manjurul Ahsan · 2025

Read A Theoretical Framework for Graph-based Digital Twins for Supply Chain Management and Optimization by Azmine Toushik Wasi, Mahfuz Ahmed Anik, Abdur Rahman, Md. Iqramul Hoque, MD Shafikul Islam, Md Manjurul Ahsan (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-03692.txt.

Introduction

Azmine Toushik Wasi and co-authors set out a framework in 2025 for running supply chains with graph-based digital twins. A digital twin is a live computer copy of a real network. A graph shows how firms, sites, and goods link to each other.

Inside, the plan has a layer that joins split data sources, a part that builds the graph, and an engine that runs tests and searches for better plans. Scores for carbon and resource use sit on the main screens, so green goals count in daily choices. The authors argue the mix can help firms spot trouble early, cut costs, grow more easily, and build supply chains that hold up better and harm nature less.

Catalog record

CREATORAzmine Toushik Wasi, Mahfuz Ahmed Anik, Abdur Rahman, Md. Iqramul Hoque, MD Shafikul Islam, Md Manjurul Ahsan
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE124 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-03692

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A Theoretical Framework for Graph-based Digital Twins for Supply Chain Management and Optimization Azmine Toushik Wasi, Mahfuz Ahmed Anik, Abdur Rahman, Md. Iqramul Hoque, MD Shafikul Islam, Md Manjurul Ahsan Abstract Abstract: Supply chain management is growing increasingly complex due to globalization, evolving market demands, and sustainability pressures, yet traditional systems struggle with fragmented data and limited analytical capabilities. Graph-based modeling offers a powerful way to capture the intricate relationships within supply chains, while Digital Twins (DTs) enable real-time monitoring and dynamic simulations. However, current implementations often face challenges related to scalability, data integration, and the lack of sustainability-focused metrics. To address these gaps, we propose a Graph-Based Digital Twin Framework for Supply Chain Optimization, which combines graph modeling with DT architecture to create a dynamic, real-time representation of supply networks. Our framework integrates a Data Integration Layer to harmonize disparate sources, a Graph Construction Module to model complex dependencies, and a Simulation and Analysis Engine for scalable optimization. Importantly, we embed sustainability metrics—such as carbon footprints and resource utilization—into operational dashboards to drive eco-efficiency. By leveraging the synergy between graph-based modeling and DTs, our approach enhances scalability, improves decision-making, and enables organizations to proactively manage disruptions, cut costs, and transition toward greener, more resilient supply chains. \abscontent 1 Introduction In today’s world of rapid globalization and technological progress, supply chains have become highly complex and interconnected, spanning multiple continents and involving various stakeholders such as manufacturers, suppliers, distributors, and customers. These stakeholders generate vast amounts of real-time data, and managing this influx is becoming increasingly difficult due to volatile market conditions, changing consumer preferences, and large-scale disruptions like natural disasters or pandemics. Organizations must navigate these challenges while striving to maintain competitive pricing, ensure timely deliveries, and build operational resilience, all under significant time and cost pressures [ 61 ] . However, modern supply chain management is not just about coordinating logistics—it involves managing intricate interdependencies across multi-level networks, where a single disruption, such as a labor strike or port congestion, can trigger widespread instability across the entire system [ 85 , 84 , 137 , 126 ] . Traditional supply chain models, which rely on linear or isolated approaches and static historical data, often fail to capture the dynamic and interconnected nature of today’s global networks [ 57 ] . While these conventional methods may support basic forecasting or retrospective analysis, they struggle to address the real-time complexities and cascading effects of modern supply chains [ 22 ] . As a result, many organizations remain reactive, responding to disruptions only after they have caused significant damage, underscoring the urgent need for more dynamic and integrated solutions in supply chain management. Graph-based modeling has emerged as a powerful approach to tackling the complexities of modern industrial setups [ 122 , 124 ] . By representing supply chains as networks, this method uses nodes to signify key entities—such as suppliers, factories, ports, and distribution centers—and edges to illustrate the connections between them, including product movements, financial transactions, and data exchanges [ 42 ] , as illustrated in Figure 1 . Unlike traditional linear models, graphs offer the flexibility to capture intricate relationships, multi-layered dependencies, and hierarchical structures that define real-world supply networks [ 40 ] . This capability is particularly important for modeling multi-tiered supplier-customer interactions, where disruptions at one level can cascade throughout the entire system. Additionally, graph-based models help supply chain managers quickly identify vulnerabilities, such as critical nodes that act as bottlenecks or single points of failure [ 42 ] . The interconnected structure of graphs also makes it possible to simulate the impact of disruptions—such as delays in raw material shipments—and assess their ripple effects on inventory, production schedules, and customer satisfaction [ 62 ] . By providing a more comprehensive and dynamic view of supply chains, graph-based modeling enables better decision-making and risk management in an increasingly uncertain global landscape. Figure 1 : Different types of relations in Supply Chain Digital Twins (DTs), originally developed by NASA to simulate spacecraft systems [ 9 ] , have become a key technology in Industry 4.0, widely used for modeling, monitoring, and optimizing physical assets [ 130 , 70 , 25 ] . A Digital Twin is essentially a virtual replica of a real-world entity—ranging from individual machines to entire supply chains—that enables continuous data exchange between the physical and digital environments. This bi-directional flow of information allows for real-time updates and closed-loop feedback, ensuring that changes in one domain are immediately reflected in the other [ 129 , 47 ] . Over time, DTs have evolved from basic “digital shadows,” where data flows only one way, into fully synchronized systems capable of real-time analytics, predictive simulations, and autonomous decision-making [ 50 , 47 , 25 ] , as demostrated in Figure 2 . Industries such as aerospace, healthcare, automotive, and smart city management have already embraced DTs for predictive maintenance and scenario planning [ 103 ] , and their impact on supply chain management is increasingly significant. DTs provide unmatched visibility into supply chain operations, tracking material flows, inventory levels, production schedules, and transportation routes using real-time data from IoT, blockchain, and cloud computing [ 158 ] . This allows organizations to proactively manage risks by forecasting disruptions such as port congestion or inventory shortages [ 142 ] . Additionally, DTs enhance predictive and prescriptive analytics, where live data simulations help anticipate and mitigate the effects of disruptions—such as delayed shipments or production halts—while also suggesting optimal response strategies to minimize downtime [ 61 ] . However, despite their advantages, implementing DTs in supply chains comes with challenges, including integrating high-quality data, ensuring security, and managing the complexity of global networks, which demand scalable architectures and advanced analytics [ 145 , 45 ] . A promising solution to these challenges is the integration of Graph Neural Networks (GNNs) with DTs. GNNs, which excel at processing graph-structured data, can help model intricate supply chain networks, detect vulnerabilities, and forecast disruptions in real time [ 5 , 52 ] . By combining DTs with GNN-driven insights, organizations can develop more resilient and data-driven supply chains, ensuring efficiency and adaptability in an increasingly complex and uncertain global landscape. Figure 2 : Digital Twins in Supply Chain Management Integrating graph-based models into Digital Twins (DTs) enhances their ability to dynamically represent entire supply chains in real time [ 150 ] . As data continuously flows in from sensors, market trends, or production updates, the graph structure within the DT adjusts, refining predictions and enabling swift responses to potential risks. This synergy is further strengthened by advanced machine learning techniques like Graph Neural Networks (GNNs), which effectively analyze both the structural connections within the network and individual node attributes, making them particularly useful for complex supply chain data [ 5 ] . By learning from historical patterns and adapting to real-time changes, GNN-powered DTs can anticipate disruptions such as production slowdowns, shipping delays, or sudden demand fluctuations [ 52 ] . Additionally, emerging research highlights that integrating blockchain technology with graph-based DTs can improve data security and traceability, which is especially critical in highly regulated industries like pharmaceuticals and aerospace, where product authenticity and compliance are essential [ 165 , 140 ]

Read the complete text