Modern agri-food supply chains require reliable and real-time data collection mechanisms to enable advanced traceability and certification, ensuring product authenticity, safety, and regulatory compliance. The growing deployment of IoT sensors, identification technologies, and embedded devices across the supply chain has significantly increased the availability of monitoring data. However, this distributed sensing paradigm also generates large volumes of heterogeneous data that exceed the processing, storage, and communication capabilities of low-cost, resource-constrained nodes. As a result, conventional cloud-centric approaches introduce latency, bandwidth overhead, and scalability limitations, hindering continuous and trustworthy traceability. This thesis addresses these challenges by proposing a framework for real-time data collection for advanced traceability and certification, specifically designed for operation at the edge of the agri-food chain. The framework enables data acquisition, preprocessing, and management directly on embedded IoT devices, reducing dependence on centralized infrastructures. To cope with the large data volume problem, on-device data reduction and dataset distillation strategies are introduced to compact raw measurements into representative, traceability-preserving information. This approach minimizes redundancy, lowers memory usage and communication costs, and enables continuous operation on microcontroller-based platforms while maintaining the integrity and usefulness of collected data. Building upon efficient edge-level data collection, the proposed system integrates secure identification and verification mechanisms, combining RFID-based product tagging, local intelligence, cloud services, and a blockchain-inspired integrity layer to create a robust binding between physical products and their digital records. This integration strengthens both traceability and anti-counterfeiting capabilities across the supply chain. The framework is validated through two representative study cases that demonstrate its feasibility and effectiveness in realistic agri-food environments. Overall, this work shows that real-time, edge-based data collection and processing constitute key enablers for scalable, low-cost, and trustworthy traceability and certification systems in modern agri-food ecosystems.
La crescente complessità delle filiere agroalimentari moderne richiede meccanismi affidabili di raccolta dati in tempo reale per supportare sistemi avanzati di tracciabilità e certificazione, garantendo autenticità del prodotto, sicurezza e conformità normativa. La diffusione di sensori IoT, tecnologie di identificazione ed embedded devices lungo la catena di approvvigionamento ha incrementato significativamente la disponibilità di dati di monitoraggio. Tuttavia, tale paradigma distribuito genera grandi volumi di dati eterogenei che superano le capacità di elaborazione, memoria e comunicazione dei nodi a basso costo. In questo contesto, approcci cloud-centrici tradizionali introducono latenze, overhead di banda e limiti di scalabilità, ostacolando una tracciabilità continua e affidabile. La tesi propone un framework per la raccolta dati in tempo reale orientato all’edge, progettato per operare direttamente su dispositivi IoT embedded nella filiera agroalimentare. Il sistema integra tecniche di riduzione dei dati e data distillation on-device per compattare le misure grezze in informazioni rappresentative e rilevanti ai fini della tracciabilità, riducendo ridondanza, costi di comunicazione e consumo di risorse. A supporto dei processi di certificazione e anti-contraffazione, il framework combina identificazione RFID, intelligenza locale e meccanismi di integrità ispirati alla blockchain, creando un legame affidabile tra prodotto fisico e registro digitale. L’approccio è validato tramite casi di studio rappresentativi in contesti agroalimentari reali.
Agri-Food Chain Data Collection for Advanced Traceability and Certification / Sebti, M.R.. - (2026 Apr 28).
Agri-Food Chain Data Collection for Advanced Traceability and Certification
Sebti, Mohamed Riad
2026-04-28
Abstract
Modern agri-food supply chains require reliable and real-time data collection mechanisms to enable advanced traceability and certification, ensuring product authenticity, safety, and regulatory compliance. The growing deployment of IoT sensors, identification technologies, and embedded devices across the supply chain has significantly increased the availability of monitoring data. However, this distributed sensing paradigm also generates large volumes of heterogeneous data that exceed the processing, storage, and communication capabilities of low-cost, resource-constrained nodes. As a result, conventional cloud-centric approaches introduce latency, bandwidth overhead, and scalability limitations, hindering continuous and trustworthy traceability. This thesis addresses these challenges by proposing a framework for real-time data collection for advanced traceability and certification, specifically designed for operation at the edge of the agri-food chain. The framework enables data acquisition, preprocessing, and management directly on embedded IoT devices, reducing dependence on centralized infrastructures. To cope with the large data volume problem, on-device data reduction and dataset distillation strategies are introduced to compact raw measurements into representative, traceability-preserving information. This approach minimizes redundancy, lowers memory usage and communication costs, and enables continuous operation on microcontroller-based platforms while maintaining the integrity and usefulness of collected data. Building upon efficient edge-level data collection, the proposed system integrates secure identification and verification mechanisms, combining RFID-based product tagging, local intelligence, cloud services, and a blockchain-inspired integrity layer to create a robust binding between physical products and their digital records. This integration strengthens both traceability and anti-counterfeiting capabilities across the supply chain. The framework is validated through two representative study cases that demonstrate its feasibility and effectiveness in realistic agri-food environments. Overall, this work shows that real-time, edge-based data collection and processing constitute key enablers for scalable, low-cost, and trustworthy traceability and certification systems in modern agri-food ecosystems.| File | Dimensione | Formato | |
|---|---|---|---|
|
PhD Thesis_Sebti Mohamed Riad_XXXVIII ciclo.pdf
accesso aperto
Descrizione: Tesi di dottorato
Tipologia:
Tesi di dottorato
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
6.62 MB
Formato
Adobe PDF
|
6.62 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


