Sovereign Data, Shared Intelligence A Polycloud Federated Learning Architecture for Privacy-Conserving Demand and Distribution Optimization in National Food Wholesale Networks
DOI:
https://doi.org/10.5281/zenodo.20442397Keywords:
Federated learning; federated analytics; big-data analytics; supply chain; privacy-preserving computation; cross-cloud; data governance; Government of Canada; transport and warehousing; food services; wholesale trade; national capital region; Amazon Web Services; Microsoft Azure; Google Cloud Platform.Abstract
National level supply chain optimization demands federated analytics across multiple sovereign clouds to respect regulatory needs and avoid privacy concerns. In traditional centralized models, neither such aspects nor the sensitivity to data latency can be suitably considered. A real-world wholesaler of the food service sector engaged in the development of demand forecasting, inventory optimization, and transportation planning models from different data domains and sources, routed through AWS, Azure, and GCP. Data sharing agreements enforced mandatory storage duration and data sharing policies to satisfy ownership rules and a business partnering model was used to coordinate application developments. State-of-the-art algorithms were applied for the federated learning building blocks, and the communication overheads associated with model exchanges were assessed.
Today's business world suffers from the lack of information about critical events that occur far away and could have a significant positive or negative influence on business outcomes. Big data opens up the possibility of having more information for analysis but brings with it new challenges and costs, especially when dealing with processing and scripting these big data analytics. The need for specialized know-how and costs are important factors that dictate the success of an analytics process and the return on investment. However, successful and careful analysis of data has its rewards. Developing scalable models on three major public clouds (AWS, Azure, and GCP) for national-level supply chain optimization and representing data and models accurately under privacy and security regulations are still in their infancy.
References
[1] Bonawitz, K., Eichner, H., Grieskamp, W., Huba, D., Ingerman, A., Ivanov, V., Kiddon, C., Konečný, J., Mazzocchi, S., McMahan, H. B., Van Overveldt, T., Petrou, D., Ramage, D., & Roselander, J. (2019). Towards federated learning at scale: System design. In Proceedings of Machine Learning and Systems (MLSys).
[2] Dai, K., et al. (2024). A privacy-preserving multi-center federated learning framework for district heating forecast. Energy and Buildings, 328, 115164.
[3] Deng, S., Zhao, H., Huang, B., Zhang, C., Chen, F., Deng, Y., Yin, J., Dustdar, S., & Zomaya, A. Y. (2024). Cloud-native computing: A survey from the perspective of services. Proceedings of the IEEE, 112(1), 12–46.
[4] Dong, H., Zhang, C., Li, G., & Zhang, H. (2024). Cloud-native databases: A survey. IEEE Transactions on Knowledge and Data Engineering, 36(12), 7772–7798.
[5] Douaioui, K., Oucheikh, R., Benmoussa, O., & Mabrouki, C. (2024). Machine learning and deep learning models for demand forecasting in supply chain management: A critical review. Applied System Innovation, 7(5), 93.
[6] Fernández, J. D., et al. (2022). Privacy-preserving federated learning for residential short-term load forecasting. Applied Energy, 314, 118633.
[7] Guo, W., et al. (2024). A comprehensive survey of federated transfer learning. Frontiers of Computer Science, 18, 183101.
[8] Hübner, N., Caspers, J., Coroamă, V. C., & Finkbeiner, M. (2024). Machine-learning-based demand forecasting against food waste: Life cycle environmental impacts and benefits of a bakery case study. Journal of Industrial Ecology, 28(5), 1117–1131.
[9] Ibrahim Khalaf, O., et al. (2024). Federated learning with hybrid differential privacy for privacy–utility trade-offs. Security and Privacy, 7(2), e374.
[10] Jauhar, S. K., et al. (2024). Explainable artificial intelligence to improve the resilience of perishable product supply chains by leveraging customer characteristics. Annals of Operations Research.
[11] Jiang, S., et al. (2024). Fed-MPS: Federated learning with local differential privacy using model parameter selection for resource-constrained CPS. Journal of Systems Architecture, 150, 103108.
[12] Kairouz, P., McMahan, H. B., Avent, B., Bellet, A., Bennis, M., Bhagoji, A. N., Bonawitz, K., Charles, Z., Cormode, G., Cummings, R., D’Oliveira, R. G. L., Eichner, H., El Rouayheb, S., Evans, D., Gardner, J., Garrett, Z., Gascón, A., Ghazi, B., Gibbons, P. B., … Zhao, S. (2021). Advances and open problems in federated learning. Foundations and Trends in Machine Learning, 14(1–2), 1–210.
[14] Khedr, A. M., & Rani, S. (2024). Enhancing supply chain management with deep learning and machine learning techniques: A review. Journal of Open Innovation: Technology, Market, and Complexity, 10(4), 100379.
[15] Kong, L., Zheng, G., & Brintrup, A. (2024). A federated machine learning approach for order-level risk prediction in supply chain financing. International Journal of Production Economics, 268, 109095.
[16] Li, J., et al. (2024). A comprehensive survey on client selection strategies in federated learning. Computer Networks, 242, 110028.
[17] Li, T., Sahu, A. K., Zaheer, M., Sanjabi, M., Talwalkar, A., & Smith, V. (2020). Federated optimization in heterogeneous networks. In Proceedings of Machine Learning and Systems (MLSys).
[18] Liu, Y., Fan, L., Mao, Y., Chen, Z., & Li, S. (2025). A survey of optimization algorithms for differential privacy in federated learning. Computer Standards & Interfaces, 95, 103___.
[19] Maher, M., Oun, O. F., Mesmeh, M. S., & El Shawi, R. (2025). FedForecaster: An automated federated learning approach for time-series forecasting. In Proceedings of the 28th International Conference on Extending Database Technology (EDBT).
[20] Mathew, C., et al. (2024). Improved data privacy with differential privacy in federated learning. Journal of Web User Interface and Usability Analytics, 15(3), 1–___.
[21] Nikkhah, A., et al. (2024). Machine learning-based life cycle assessment for improving food supply chains toward sustainability. Sustainable Production and Consumption, 49, 1–___.
[22] Perifanis, V., Pavlidis, N., Koutsiamanis, R.-A., & Efraimidis, P. S. (2023). Federated learning for 5G base station traffic forecasting. Computer Networks, 227, 109950.
[23] Ranpara, R., et al. (2025). A semantic and ontology-based framework for enhancing interoperability and automation in IoT systems. Discover Internet of Things, 5, 122.
[24] Shadid, N., et al. (2025). A systematic review of data-driven approaches to food demand forecasting and waste reduction. Sustainable Futures, 7, 100___.
[25] Shan, F., et al. (2025). A survey of optimization algorithms for differential privacy in federated learning. Computer Standards & Interfaces, 95, 103___.
[26] Shi, C., et al. (2025). FedAWA: Adaptive optimization of aggregation weights in federated learning using client vectors. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).
[27] Stranieri, F., et al. (2024). Combining deep reinforcement learning and multi-stage stochastic programming to address the supply chain inventory management problem. International Journal of Production Economics, 268, 109099.
[28] Tayyeh, H. K., et al. (2024). A differential privacy approach in federated learning. Computers, 13(11), 277.
[29] Thwal, C. M., et al. (2024). Transformers with attentive federated aggregation for time-series forecasting in non-IID settings. Expert Systems with Applications, 252, 123___.
[30] Wan, X., Yang, D., Wang, T., & Deveci, M. (2023). Closed-loop supply chain decision considering information reliability and security: Should the supply chain adopt federated learning decision support systems? Annals of Operations Research.
[31] Wei, J., et al. (2024). FedTWA: A federated learning model for supply chain demand prediction with temporal pattern attention. In Proceedings of the ACM International Conference on AI in Finance / Applied AI (ACM).
[32] Xia, J., et al. (2024). PT-ADP: A personalized privacy-preserving federated learning scheme based on transaction mechanism. Information Sciences, 669, 120519.
[33] Yang, F., et al. (2024). An explainable federated learning and blockchain-based secure credit modeling method. European Journal of Operational Research, 317(2), 449–467.
[34] Zheng, G., Kong, L., & Brintrup, A. (2023). Federated machine learning for privacy preserving, collective supply chain risk prediction. International Journal of Production Research, 61(23), 8115–8132.
[35] Zheng, G., Ivanov, D., & Brintrup, A. (2025). An adaptive federated learning system for information sharing in supply chains. International Journal of Production Research, 63(11), 3938–3960.
[36] Zogaan, W. A., et al. (2025). Leveraging deep learning for risk prediction and resilience in supply chains: A big data perspective. Journal of Big Data, 12, 143.
Additional Files
Published
Data Availability Statement
The paper notes that public datasets from the Kaggle Data Exchange were used for demand forecasting, and that confidential data for remaining modeling domains would be generated with a data generator. Actual sales and logistics data from a national food distributor were used for inventory optimization validation, but no formal data availability statement or repository link is provided.
