Securing the Food Supply Chain Resilience – The SecureFood Project

Securing the Food Supply Chain Resilience – The SecureFood Project Sofia Kokonezi1, Dr. Angelos Amditis1, Giannis Kanellopoulos1, Konstantinos Touloumis2, Apostolos Kapetanios3, Dr. Dimitris Drakoulis4 and Dr. Georgios Kolionis5 1. ISENSE Group-Institute of Communication and Computer Systems (ICCS), Athens, Greece 2. European Dynamics SA, Athens, Greece 3. European Dynamics Luxembourg SA, Luxembourg, Luxembourg 4. InnovActs, Nicosia, Cyprus 5. EXUS, Athens, Greece Corresponding author: sofia[email protected] Abstract: The increasing complexity and vulnerability of food supply chains demand integrated, data-driven solutions to ensure resilience and adaptability. In response, the EUfunded SecureFood project introduces a digital ecosystem composed of interoperable tools designed to enhance logistics, transparency, and proactive decision-making across all stages of the food system. This paper presents the SecureFood architecture and the functionalities of its five key tools: the Observatory Dashboard for real-time monitoring of essential metrics and access to system components; an AI-powered Early Warning System for risk prediction; the Digital Twin for scenario simulation and vulnerability assessment; the RESILOG tool for resilient route planning; and a blockchain-based Information Exchange Platform for secure data sharing. These tools are built to operate collaboratively, under a common authentication system, and support modular, scalable deployment in line with the Physical Internet paradigm. Tool interactions and functionalities will be demonstrated through four case studies—covering grain, dairy, aquaculture, and fruit and vegetable supply chains—highlighting their value in enhancing situational awareness, stakeholder coordination, and operational continuity. The findings underscore the potential of SecureFood to drive the transition toward hyperconnected, intelligent, and sustainable food logistics systems. Keywords: Supply Chain Resilience, Digital Twin technology, Physical Internet, Predictive Analytics, Block-chain-Based Information Exchange, Resilient Route Planning, Early Warning System, Data Visualisation. Physical Internet (PI) Roadmap Fitness: Select the most relevant area(s) for your paper according to the PI roadmaps adopted in Europe and Japan: ☐ PI Nodes (Customer Interfaces, Logistic Hubs, Deployment Centers, Factories), ☐ Transportation Equipment, ☒ PI Networks, ☒ System of Logistics Networks, ☐ Vertical Supply Consolidation, ☐ Horizontal Supply Chain Alignment, ☒ Logistics/Commercial Data Platform, ☐ Access and Adoption, ☐ Governance. Targeted Delivery Mode-s: ☒ Paper, ☐ Poster, ☐ Flash Video, ☒ In-Person presentation 1 Introduction The global food supply chain is a highly complex and interconnected system that ensures the movement of food from production to consumption. However, it faces numerous challenges Securing the Food Supply Chain Resilience – The SecureFood Project Research Contribution 2 that threaten its efficiency, resilience, and sustainability. One of the most critical issues is logistics inefficiencies, which lead to delays, increased costs, and food waste. The food supply chain requires well-coordinated transportation and storage systems, yet disruptions caused by climate change, geopolitical instability, and labour shortages frequently hinder operations (Rasshyvalov et al., 2024). In addition, the COVID-19 pandemic further exposed vulnerabilities in food supply chains by causing massive disruptions in logistics, trade restrictions and labour availability. Lockdowns and border closures disrupted transportation networks, leading to bottlenecks in food distribution and shortages of essential goods (Li et al., 2023). The pandemic also highlighted the need for more flexible and adaptive supply chain models that can withstand sudden shocks and disruptions (Ivanov and Dolgui, 2020; Aday S. and Aday M. 2020). Inadequate route planning and inefficient distribution networks further exacerbate these problems, causing significant losses, particularly in perishable goods (Fatorachian et al., 2025). Another major challenge is fragmented information exchange across stakeholders. Food supply chains involve multiple actors -farmers, producers, processors, distributors, and retailerseach operating with different systems and data sources. The lack of real-time, standardized datasharing mechanisms creates inefficiencies and impede collaboration (Akinbamini et al., 2025). Blockchain and digital platforms offer promising solutions (Abideen et al., 2021; Li et al., 2021), yet significant challenges that need to be overcome still exist, with the most prominent being interoperability issues and concerns about data security (Li et al., 2021). Early warning systems (EWS) are also critical in mitigating risks associated with food supply chain disruptions, such as food contamination, extreme weather events, and logistics bottlenecks. However, data-driven decision-making in food logistics remains underutilized. Many organizations still rely on reactive rather than proactive strategies, missing opportunities to optimise supply chain visibility and performance through predictive analytics and artificial intelligence (Adewale and Ahsan, 2025). Addressing these challenges requires innovative digital solutions that optimise logistics, enhance information transparency, and enable proactive, data-driven decision-making. In response, the EU-funded SecureFood project has developed a comprehensive ecosystem designed to strengthen the resilience and sustainability of food supply networks. At its core, this ecosystem integrates cutting-edge digital tools, including the Digital Twin for smart supply chains, an Early Warning System for risk detection, the Resilient Route Planning tool for optimising transport logistics, a blockchain-based Information Exchange Platform to ensure secure and transparent data sharing and the Observatory Dashboard, which serves as the system’s central access point and provides monitoring and visualization of key resilience metrics. Together, these technologies streamline operations, mitigate disruptions, and enhance the overall efficiency of food supply chain management. 2 System Layers The reference architecture of the SecureFood project is structured across four main layers: Business, Process, Technology, and Data. Together, these layers create a resilient, interoperable, and adaptive digital ecosystem that supports food supply chain stability. While SecureFood includes a range of innovative digital tools, this paper focuses on those most directly aligned with the Physical Internet vision - namely the Observatory Dashboard (OD), Early Warning System (EWS), Digital Twin (DT), Information Exchange Platform (IEP), and the Resilient Route Planning tool (RESILOG). The Business Layer defines the overarching strategic objectives, priorities, and governance mechanisms that shape resilient food systems. This layer brings together key stakeholders - Securing the Food Supply Chain Resilience – The SecureFood Project Research Contribution 3 including producers, processors, logistics operators, and policymakers - to coordinate responses to food security risks and align with EU goals such as sustainability, decarbonization, and supply chain transparency. Decision-making at this level is supported by the risk assessment model embedded within the Digital Twin for Smart Supply Chain, which enables strategic scenario planning and vulnerability evaluation across different stages of the food supply network. The Process Layer encapsulates the workflow and operations necessary to implement resilience strategies. Key SecureFood tools such as RESILOG, the DT, the EWS, and the OD, enable predictive and proactive action. RESILOG and the DT allow route optimisation, scenario-based planning and vulnerability analysis. The EWS provides AI-driven early detection of risks, and the OD ensures real-time monitoring and visualization of risk indicators. These tools are especially relevant to the Physical Internet paradigm, promoting hyperconnected, decentralized logistics decision-making. The Technology Layer provides the enabling infrastructure for secure, scalable, and interoperable deployment of digital tools. Cloud-native architectures, containerised microservices, and secure APIs allow for real-time integration across internal modules and with external datasets. The IEP, built on blockchain, exemplifies trust-by-design infrastructure by ensuring traceability and immutability of shared data. Secure authentication systems (e.g., Keycloak, DIDs) ensure controlled data access. Together, these technologies underpin Physical Internet principles such as modularity, decentralization, and transparency. This foundational Data layer manages the collection, processing, and storage of supply chain data. It integrates real-time feeds (e.g., weather, transport delays, stock levels), historical performance data, and open-access datasets from publicly available sources like Copernicus and FAO. Advanced analytics and machine learning models transform raw data into predictive insights, supporting tools such as the EWS and DT. The IEP and OD further ensure structured data exchange and visualization, empowering stakeholders to act on a shared understanding of risks and conditions. In the context of the Physical Internet, this data backbone is essential for synchronised, data-driven logistics optimisation across a distributed network. By integrating these four layers, SecureFood delivers a comprehensive digital architecture to enhance food supply chain resilience. The tools emphasized in this paper provide practical, PIaligned solutions to today’s pressing logistics challenges - ensuring that food flows remain efficient, adaptive, and secure in an increasingly complex global environment. 3 Digital Tools 3.1 Observatory Dashboard The Observatory Dashboard (OD) is a unified digital platform that facilitates the monitoring, analysis, and visualization of essential metrics pertaining to food system resilience and security. By aggregating and visualizing data from diverse sources, including weather forecasts, food commodity prices, demand fluctuations, and production data, the OD enhances situational awareness and supports evidence-based decision-making. Through its secure API integration and in cooperation with blockchain-supported information exchange and the EWS, it fosters an interconnected network, aligning with the Physical Internet's vision of open and efficient logistics. The OD incorporates cutting-edge web technologies to provide an intuitive and responsive interface for users. Developed using React and TypeScript, the system ensures scalability, maintainability, and high-performance rendering of real-time data. Interactive features such as Securing the Food Supply Chain Resilience – The SecureFood Project Research Contribution 4 dynamic charts and historical trend analysis allow stakeholders to track disruptions and adjust their operations proactively. Secure access is managed via Keycloak authentication, ensuring compliance with industry standards and safeguarding sensitive data. Additionally, React Query and Zustand optimise state management, while WebSockets or Server-Sent Events (SSE) enable real-time updates, a crucial feature for logistics coordination in Physical Internetenabled networks. Beyond its technical framework, the OD is designed to facilitate collaboration across food supply chain actors. By standardizing and sharing key indicators, it promotes transparency in food markets, improves risk assessment, and enhances supply chain interoperability - a core principle of the Physical Internet. Its ability to provide clear, analytical tables and dashboards empowers food supply chain actors such as transport operators and wholesalers to optimise routing and inventory management based on real-time conditions. As global food systems become increasingly interconnected and digitalized, the Observatory Dashboard serves as a critical enabler of resilient, data-driven, and hyperconnected logistics, paving the way for a more efficient and sustainable food supply chain ecosystem. 3.2 Early Warning System The Early Warning System (EWS) is an advanced module which suggests and predicts risks relating to different food supply chain stages, to increase risk awareness, issue prevention and support decision making for stakeholders. The EWS receives input from both internal sources (other SecureFood components, e.g. the IEP) and external data sources (e.g. Copernicus, Eurostat) and uses AI risk identification and prioritization techniques along with real-time data analytics in conjunction with user evaluation to predict possible risks along with a criticality level personalized to each user according to their risk profiles. The estimated risks are transferred as alerting messages to other SecureFood components (e.g. the DT to perform simulations) and displayed in the EWS Graphical User Interface (GUI) with their details. The GUI also assists users in monitoring and managing the estimated risks during their entire lifespan.The EWS architecture comprises of several key components and is depicted in the figure below: Figure 1: EWS Architecture Securing the Food Supply Chain Resilience – The SecureFood Project Research Contribution 5 • Graphical User Interface (GUI): Enables users (transport operators, producers, wholesalers, policy makers) to input food supply chain parameters, define their risk thresholds, and receive automated alerts. • AI Layer: Uses AI and Multiple Criteria Decision Analysis (MCDA) techniques to estimate risks based on user-defined profiles. • REST APIs: Establish endpoints for secure interactions of external systems with the EWS. • Risk Identification Module: Identifies risks according to user risk profiles, each risk is followed with a criticality message depicting the urgency of actions needed to be taken by the user. • Data Acquisition Pipelines: Automates the collection, processing, and storage of realtime data from diverse sources. • Data Storage: Maintains structured datasets, including user-defined parameters, external and internal data sources, and risk estimations. • Identify Access Management Layer (vertical): Ensures secure and role-based access to system resources and risk profiles. It authenticates users and services, manages credentials, and enforces permissions across all system layers. By implementing this modular and interoperable architecture, the EWS enhances the predictive capabilities of food supply chain actors, allowing them to: • Predict extreme weather conditions droughts/snowstorms (based on temperature, windspeed, precipitation). • Predict extreme sea water conditions (temperature, chlorophyll-phytoplankton, oxygen) • Identify processing risks (increased energy costs, prices). • Generate warnings for transportation disruptions (e.g. accidents). The EWS is deployed using a Windows/Ubuntu Virtual Machine (VM) infrastructure with Docker containers, ensuring scalability, security, and seamless integration within existing digital ecosystems. It utilizes: • Frontend (GUI): Angular-Ionic for a responsive user experience. • Risk Estimation APIs: Java Spring Boot, ensuring efficient risk computation. • Data Storage: PostgreSQL, offering high-performance risk data management. In alignment with Physical Internet principles, the EWS functions as a decentralized, interconnected risk prediction platform and alerting system. It enables food supply chain actors to share real-time alerts and optimise logistics and resource allocation dynamically. By embedding data-driven intelligence into supply chain processes, it contributes to hyperconnected, autonomous decision-making, strengthening food security and resilience in an increasingly volatile global landscape. 3.3 Digital twin-driven smart supply chain The Digital Twin-driven Smart Supply Chain (DT) is a virtual replica regarding production quantities and disruptions in food supply chain, designed to enhance situational awareness, improve decision-making, and ensure food supply chain resilience. By integrating historical and real-time data with predictive analytics, simulation and risk assessment models, the tool enables stakeholders to anticipate disruptions, explore what-if scenarios, and optimise supply chain performance dynamically. It serves as a strategic solution to improve operational visibility, test potential impact of disruptions and response strategies, and assess vulnerabilities Securing the Food Supply Chain Resilience – The SecureFood Project Research Contribution 6 across different stages of the food supply chain - while aligning with SecureFood’s overarching objectives for food security and sustainability. This web-based tool is designed to act as both a monitoring tool and a predictive engine for identifying turnaround scenarios in the face of potential disruptions. It incorporates supervised machine learning algorithms - such as Random Forest and Decision Trees - to forecast disruption probabilities across the supply chain, as well as advanced mathematical models to assess risk exposure and vulnerabilities. These models process a combination of historical and real-time data to simulate system behaviour under stress, helping stakeholders anticipate disruptions and formulate optimal intervention strategies. By modeling both internal operational factors and external influences, the tool delivers a comprehensive view of supply chain stability. Through techniques like Mixed-Integer Linear Programming, the system supports data-driven decision-making aimed at strengthening the core pillars of food security: availability, access, utilization, and stability. Embedded within a broader suite of SecureFood components - including the Early Warning System, the Information Exchange Platform, and RESILOG - the Digital Twin facilitates dynamic data sharing and coordinated risk management. From a technological perspective, the system architecture includes containerized back-end and front-end layers, utilizing scalable, cloud-based infrastructure. A modular design supports simulation engines, rendering modules, and a data orchestration service, enabling input collection, execution of AI and mathematical models, and visualization through a user-friendly interface. Alerts and reports are generated in standardized formats, facilitating quick dissemination of actionable insights. The Digital Twin aligns with the principles of the Physical Internet by enabling modular, interoperable, and real-time orchestration of food logistics networks. It promotes standardized data exchange, decentralized decision support, and dynamic reconfiguration of assets and flows in response to changing conditions - key enablers of hyper-connected, efficient, and resilient supply chain ecosystems. 3.4 Information Exchange Platform The Information Exchange Platform (IEP) is a digital tool designed to enhance communication, transparency, and trust among stakeholders across the food supply chain - from producers and retailers to logistics actors and public authorities at national and European levels. Its core purpose is to enable secure, structured, and timely information sharing through a decentralised architecture, ensuring all participants can access and act on reliable data. By leveraging blockchain technology, the platform guarantees data integrity, traceability, and trust across the ecosystem - key enablers of transparency and accountability. IEP is composed of three key modules: • General Information Exchange Module: Facilitates collaborative knowledge sharing by enabling end-users (food actors, transport operators, retailers and authorities) to exchange general updates, circulate best practices, and engage in structured communication. This promotes openness and cooperation throughout the food supply chain system. • Stock Reporting Module: Provides food actors with a secure mechanism to notify authorities of available commodity stocks using standardized forms. This ensures an accurate, real-time view of food availability, which is essential for informed policymaking and rapid response during crises. Securing the Food Supply Chain Resilience – The SecureFood Project Research Contribution 7 • Incident Reporting Module: Enables the structured and timely reporting of incidents, alerting both authorities and other interdependent actors within the supply chain. This supports early detection, response coordination, and mitigation of cascading effects. Each module collects input directly from users through predefined forms that ensure consistency, completeness, and harmonization of data. The submitted information is visible to other authorized users and can also be shared with other SecureFood tools through a dedicated API - fostering interoperability and integration across the broader digital ecosystem. The architecture of the IEP is designed around a decentralized application (D-App) model, structured into six functional layers. Three of these form the blockchain core, while the remaining layers ensure seamless interaction with users and other systems. The diagram below shows the Information Exchange Platform’s architecture which consists of: Figure 2: IEP architecture • Distributed Network & Credentials: Defines user identities using a combination of centralized credentials and Distributed Identifiers (DIDs), ensuring secure access and user authentication. It includes public and private keys in order for users to submit a transaction in the blockchain. An external smart wallet application (e.g., Metamask) manages user keys and enables secure blockchain transactions. • Distributed Ledger: Serves as the immutable data repository and transaction engine, operating on Ethereum blockchain technology with a Proof-of-Authority (PoA) consensus mechanism to validate and secure entries. • Distributed Processing: Defines the methods of data execution to be performed within the general blockchain distributed network. These methods are imprinted in a smart contract and define all actions performed by the Application Binary Interface (ABI), while corresponding with the ledger to successfully submit the transaction. For IEP, the smart contract will imprint all methods of data execution needed for the three modules to be properly performed by the users. To provide a user-friendly experience, the blockchain back-end is connected to: • Web3 Library Layer: Bridges smart contracts with the application layer, abstracting complexity and enabling secure interactions. • Backend API: Enables communication between the blockchain and front-end applications, while also providing endpoints for other SecureFood tools—such as the Securing the Food Supply Chain Resilience – The SecureFood Project Research Contribution 8 Early Warning System or the Observatory Dashboard - to retrieve alerts or stock updates. • Graphical User Interface (GUI): Offers a clean, intuitive interface for end-users to submit information, report incidents or commodities stocks, and interact with other actors in the platform. Access can be initiated through the SecureFood Observatory Dashboard, provided identity mapping with DIDs is in place. The Information Exchange Platform acts as a digital enabler of trust and collaboration, providing the infrastructure needed for synchronized, adaptive, and efficient food supply chain operations—precisely the type of integration envisioned by the Physical Internet paradigm. 3.5 Resilient Route Planning tool The RESILOG is a cloud-based, data-driven platform designed to foster digital and horizontal collaboration between shippers, carriers, and other multimodal logistics stakeholders. Its core function is the identification of the best available transport options, considering both singlemode and multimodal alternatives. Acting as a collaborative digital marketplace, RESILOG facilitates synchromodal transport planning, by intelligently matching transport demand and supply. It also uncovers opportunities for cargo consolidation and bundling opportunities across different shippers and logistics service providers, enhancing efficiency and flexibility across the supply chain. RESILOG offers key functionalities such as: • Mapping commercially viable transport routes between two points, incorporating mode combinations and logistics constraints • Forecasting route performance based on historical and disruption-related data • Increasing load factors for transport vehicles and standardized modular units • Reducing empty returns and overall transport costs While RESILOG can function independently, it also includes an API integration layer for seamless data exchange with other tools in the SecureFood ecosystem - such as the Digital Twin, the Information Exchange Platform and Early Warning System - enabling coordinated responses and improved situational awareness. Its flexible design allows it to operate in both federated and standalone modes: • As part of a federation, using the authentication and authorisation controls defined by the federation, encouraging data sharing and interoperability. • As a standalone platform, using its own user authentication and authorisation controls to meet varying user needs. More specifically, to effectively support decision-making in both long-term transport planning and ad-hoc rerouting after disruptions, RESILOG offers a user-centric interface that enables logistics stakeholders to input key parameters and execute scenario-based route planning. Complementing this, an Application Programming Interface (API) facilitates seamless bulk data exchange with users’ back-office systems, allowing the import of operational data and the provision of up-to-date transport schedules and available capacities. The API also supports realtime updates, ensuring the tool reflects the most current network conditions. At its core, RESILOG leverages advanced graph-based technology to dynamically explore all feasible multimodal transport routes between two locations, breaking them down by leg and providing key indicators such as turnaround time, estimated CO₂ emissions, and cost. Enhancing this capability, a predictive algorithm utilizes historical data and disruption-related parameters to Securing the Food Supply Chain Resilience – The SecureFood Project Research Contribution 9 assess and forecast route availability, delivering proactive insights that strengthen transport resilience and continuity in the face of evolving supply chain risks. RESILOG embodies the principles of the Physical Internet (PI) by enabling open, modular, and intelligent logistics planning. It encourages collaboration across traditionally siloed stakeholders, standardises data sharing, and supports interconnected routing strategies that mirror PI’s vision of seamlessly moving goods through a shared, responsive network. By facilitating the real-time reconfiguration of logistics flows based on disruptions, RESILOG contributes to a more adaptive, transparent, and sustainable food transport system, aligning logistics infrastructure with Physical Internet values. 4 SecureFood System: Tools’ Interactions and Case Study Applications 4.1 Tool Interactions The SecureFood system is built upon the safe and seamless integration of multiple digital tools, working together to support food system resilience and security. Central to this architecture is a common authentication mechanism, which ensures secure and role-based access across all SecureFood tools. At the core of the user experience is the Observatory Dashboard (OD), which serves as the main entry point to the SecureFood system. Through the OD, users can visualize key metrics related to food system resilience and disruptions. It also provides direct links to access the graphical user interfaces (GUIs) of the other SecureFood tools, enabling smooth navigation and operational efficiency. The Information Exchange Platform (IEP) functions as a coordination hub for incident reporting and communication. It sends confirmed incidents to the EWS to be used as input for the EWS’s multi-criteria analysis and end-user notification system. Simultaneously, the IEP also communicates with the Digital Twin (DT) by transmitting identified disruptions, enriching the DT’s data sources. Additionally, it provides the EWS with production level information, triggering incident alerts when thresholds defined by user profiles are exceeded. RESILOG, the resilient route logistics optimisation tool, can be accessed either independently or through the Observatory Dashboard, giving flexibility to transport stakeholders in identifying optimal routes and managing disruptions in delivery. Through this interconnected framework, SecureFood enables data-driven, anticipatory decision-making, ensuring that stakeholders can act swiftly and effectively to maintain food system stability. 4.2 Case Studies The SecureFood system and its integrated digital tools will be demonstrated and validated through four distinct case studies, each reflecting different food supply chains, regional contexts, and resilience challenges across Europe. These scenarios are selected to ensure a comprehensive evaluation of the system’s capabilities in diverse settings, highlighting how SecureFood supports risk-informed decisions, operational coordination, and supply chain resilience. • Grain Case Study – Ukraine • Fruits and Vegetables – Portugal • Fish and Aquaculture – Greece and Belgium • Milk and Dairy Products – Greece and Finland