Optimizing Edge-Cloud RAG for Industrial IoT with On-Prem Vector Cache Coherency

Aima Ajmal; M. Zulkifil Hassan

Authors

Aima Ajmal University of the Central Punjab, Lahore, Pakistan
M. Zulkifil Hassan University of the Central Punjab, Lahore, Pakistan

Keywords:

Industrial IoT, Edge Computing, Cloud Computing, Retrieval-Augmented Generation, Vector Cache

Abstract

Industrial Internet of Things (IIoT) is revolutionizing the world of industries globally with the ability to exchange data in real-time between connected devices. However, the traditional approach of using the cloud alone to achieve this is facing a major challenge in terms of latency and security issues. This study aims to introduce a new hybrid approach of Edge-Cloud Retrieval Augmented Generation (RAG) with the inclusion of a Localized Vector Cache Coherency Protocol to overcome the challenges of the traditional approach of using the cloud alone in IIoT systems. The proposed system implements an edge tier for local vector storage and a cloud tier for global vector storage with the use of a multi-level synchronization algorithm to ensure data consistency across IIoT nodes. From the test results conducted on the system, the hybrid approach of using the Edge-Cloud Retrieval Augmented Generation with the Localized Vector Cache Coherency Protocol achieved a vector retrieval accuracy of 96%, while at the same time increasing the cache hit rate to 90% compared to the traditional approach of using the cloud alone. Moreover, the implementation of the multi-level vector cache system resulted in a significant reduction of the average end-to-end latency of the IIoT system from 350 ms with the traditional approach of using the cloud alone to 95 ms with the hybrid approach of using the Edge-Cloud Retrieval Augmented Generation with the Localized Vector Cache Coherency Protocol.

References

“Robust And Explainable Hybrid Deep Learning Model For Real-Time Zero-Delay Botnet Detection In Industrial Iot.” Accessed: Mar. 18, 2026. [Online]. Available: https://www.researchgate.net/publication/395005405_Robust_And_Explainable_Hybrid_Deep_Learning_Model_For_Real-Time_Zero-Delay_Botnet_Detection_In_

Industrial_Iot

“A Survey on Edge Computing (Ec) Security Challenges: Classification, Threats, and Mitigation Strategies | Request PDF.” Accessed: Apr. 03, 2026. [Online]. Available: https://www.researchgate.net/publication/389179422_A_Survey_on_Edge_Computing_Ec_Security_Challenges_Classification_Threats_and_Mitigation_Strategies

Alfredo Barron, Dante D. Sanchez-Gallegos, “On the Efficient Delivery and Storage of IoT Data in Edge–Fog–Cloud Environments,” Sensors, vol. 22, no. 18, p. 7016, 2022, doi: https://doi.org/10.3390/s22187016.

Jamilu Ibrahim Argungu, Mustapha Malami Idina, “A Survey of Edge Computing Approaches in Smart Factory,” IJRCCE, vol. 12, no. 9, pp. 17–30, 2023, doi: 10.17148/IJARCCE.2023.12903.

S. Kumar, P. Singh, and A. Singh, “A Review of Optimized Computational Strategies for IoT: Cloud, Fog, and Edge Computing Approaches,” Proc. 5th Int. Conf. Pervasive Comput. Soc. Networking, ICPCSN 2025, pp. 925–930, 2025, doi: 10.1109/ICPCSN65854.2025.11035529.

Horia Alexandru Modran, “Leveraging RAG with ACP & MCP for Adaptive Intelligent Tutoring,” Appl. Sci., vol. 15, no. 21, p. 11443, 2025, doi: https://doi.org/10.3390/app152111443.

Heorhii Kuchuk, Eduard Malokhvii, “Integration of IoT with Cloud, Fog, and Edge Computing: a Review,” Adv. Inf. Syst., vol. 8, no. 2, pp. 65–78, 2024, doi: 10.20998/2522-9052.2024.2.08.

Chao Jin, Zili Zhang, Xuanlin Jiang, Fangyue Liu, Xin Liu, Xuanzhe Liu, Xin Jin, “RAGCache: Efficient Knowledge Caching for Retrieval-Augmented Generation,” arXiv:2404.12457, 2024, [Online]. Available: https://arxiv.org/abs/2404.12457

“(PDF) Adaptive Q-Learning for Fair and Dynamic Server Selection in Edge Computing: Addressing Latency Variability in Real-Time Applications Chronicle Abstract.” Accessed: Mar. 18, 2026. [Online]. Available: https://www.researchgate.net/publication/388122520_Adaptive_Q-Learning_for_Fair_and_Dynamic_Server_Selection_in_Edge_Computing_Addressing_Latency_Variability_in_Real-Time_Applications_Chronicle_Abstract

P. Danvirutai, S. Tola, B. Yuangsoi, S. Charoenwattanasak, K. Thaiso, and C. Srichan, “Aquaculture Automation using Retrieval-Augmented Generation with Large Language Model and AIoT Systems,” 2025 5th Int. Conf. Comput. Commun. Inf. Syst., pp. 143–147, Feb. 2025, doi: 10.1109/CCCIS64581.2025.00032.

Deafallah Alsadie, “A Comprehensive Review of AI Techniques for Resource Management in Fog Computing: Trends, Challenges and Future Directions,” IEEE Access, vol. 12, 2024, doi: 10.1109/ACCESS.2024.3447097.

Taha Hasnain Raza, Adnan Ahmad, “Enhancing Multivariate Data Classification Using Graph Convolutional Networks: A Comparative Evaluation with PCA and t-SNE,” VAWKUM Trans. Comput. Sci., vol. 13, no. 2, 2025, [Online]. Available: https://vfast.org/journals/index.php/VTCS/article/view/2052

“A Comprehensive Review On IoT And Edge Computing In Electronics: Trends, Challenges, And Future Directions » Article.” Accessed: Apr. 04, 2026. [Online]. Available: https://journals.stmjournals.com/article/article=2026/view=238860/

“(PDF) Federated Learning for Distributed AnomalyDetection in Network Traffic Using GRU-BasedModels.” Accessed: Mar. 18, 2026. [Online]. Available: https://www.researchgate.net/publication/390371660_Federated_Learning_for_Distributed_AnomalyDetection_in_Network_Traffic_Using_GRU-BasedModels

Q. He et al., “Integrating IoT and 6G: Applications of Edge Intelligence, Challenges, and Future Directions,” IEEE Trans. Serv. Comput., vol. 18, no. 04, pp. 2471–2488, Jul. 2025, doi: 10.1109/TSC.2025.3586152.

R. R. Nikam and D. Motwani, “Towards Decentralized Fog Computing: A Comprehensive Review of Models, Architectures, and Services,” Lect. Notes Networks Syst., vol. 818, pp. 135–147, 2024, doi: 10.1007/978-981-99-7862-5_11.

“(PDF) Advancing Computational Efficiency: Innovations in Parallel and Distributed Applications and Algorithms.” Accessed: Mar. 18, 2026. [Online]. Available: https://www.researchgate.net/publication/391430816_Advancing_Computational_Efficiency_Innovations_in_Parallel_and_Distributed_Applications_and_Algorithms

M. Z. Hussain et al., “IoT Data Management and A Brief Analysis of IoT in the Health Industry,” 2023 Comput. Appl. Technol. Solut. CATS 2023, 2023, doi: 10.1109/CATS58046.2023.10424074.

G. K. Walia, M. Kumar, and S. S. Gill, “AI-Empowered Fog/Edge Resource Management for IoT Applications: A Comprehensive Review, Research Challenges, and Future Perspectives,” IEEE Commun. Surv. Tutorials, vol. 26, no. 1, pp. 619–669, 2024, doi: 10.1109/COMST.2023.3338015.

Corizon Sinar Arainy, “The Role of Edge Computing in Secure and Scalable IoT Systems: A Global Perspective,” Digit. J. Comput. Sci. Appl., vol. 3, no. 1, 2025, [Online]. Available: https://journal.idscipub.com/index.php/digitus/article/

view/856

“Artificial Intelligence Based Smart and Secured Applications: Third International Conference, ASCIS 2024, Rajkot, India, October 16–18, 2024, Revised Selected Papers, Part V | Springer Nature Link.” Accessed: Mar. 18, 2026. [Online]. Available: https://link.springer.com/book/10.1007/978-3-031-86302-8

M. Z. Hasan et al., “Urban Data Management using Cloud Computing and IoT,” pp. 1–12, Feb. 2024, doi: 10.1109/cats58046.2023.10424408.

“A Hybrid Lightweight And Explainable Federated Learning Model For Real-Time Intrusion Detection In Resource-Constrained Iot Environments.” Accessed: Mar. 18, 2026. [Online]. Available: https://www.researchgate.net/publication/392778520_A_

Hybrid_Lightweight_And_Explainable_Federated_Learning_Model_For_Real-Time_Intrusion_Detection_In_Resource-Constrained_Iot_Environments

Muhammad Zunnurain Hussain, Zurina Mohd Hanapi, “Low Network Power Challenges in IoT-Based Applications in Smart Cities,” J. Adv. Res. Appl. Sci. Eng. Technol., 2024, doi: 10.37934/araset.54.2.218237.

Brian Sal, Diego García-Saiz, “Domain-specific languages for the automated generation of datasets for industry 4.0 applications,” J. Ind. Inf. Integr., vol. 41, 2024, [Online]. Available: https://www.sciencedirect.com/science/article/pii/

S2452414X24001018

M. A. Bin Afarin, M. A. Bin Isa, S. Z. M. Hashim, H. N. A. Hamed, and A. Matthew, “Mapping the Intellectual Landscape Of Retrieval-Augmented Generation (RAG): A Bibliometric Analysis,” KSII Trans. Internet Inf. Syst., vol. 20, no. 1, pp. 284–307, Jan. 2026, doi: 10.3837/TIIS.2026.01.013.