Citation :

A. Kumaraswamy, G. Shailaja, Tanmaya Bhoi, Prashant Sharma, T.B Sivakumar, Umang Gupta, "Agentic AI Controller for Autonomous Industrial Network Recovery," International Journal of Electrical and Electronics Engineering, vol. 13, no. 9, pp. 43-54, 2026. Crossref, https://doi.org/10.14445/23488379/IJEEE-V13I9P104

Abstract

Industrial networks must ensure that there is a path to re-establish communications and recover from incident failures, congestion, communication failures, and controller outages without violating process deadlines or security segmentation. This paper presents AAC-AINR, a new agentic AI controller that enables topology-aware diagnosis, retrieval-grounded planning, digital-twin validation, transactional Software-Defined Networking (SDN) execution and constrained proximal policy optimization. The controller divides up the monitoring, diagnosis, planning, validation, execution and learning into the bounded agents, and it does not permit the language model to directly control the devices with arbitrary instructions, but instead sends instructions to the policy. Industrial attack evidence is replayed from 25-400 device industrial SDN topologies, which is used in experiments, in order to detect the attack. AAC-AINR achieves a median recovery time of 29.8s in a 100-device environment, recovers 94.6% of the throughput and minimizes residual packet loss to 2.3%. The proposed approach is also found to be successful with a device count of 400 in the simulation with recovery rate of 95.1%. The ablation results show that it is important to validate the digital twin otherwise unsafe plans will be implemented, while the graph diagnosis and case memory minimize unwanted changes. The framework will be a supervisory recovery layer to be used in conjunction with a deterministic industrial failover mechanism.

Keywords

Agentic AI, Autonomous recovery, Industrial networks, SDN, Digital twin, PPO, Multi-agent control, Cyber resilience.

References

1. Anees Ur Rehman, Rui L. Aguiar, and João P. Barraca, “Fault-Tolerance in the Scope of Software-Defined Networking,” IEEE Access, vol. 7, pp. 124474-124490, 2019.
   [CrossRef] [Google Scholar] [Publisher Link]
2. Werner Kritzinger et al., “Digital Twin in Manufacturing: A Categorical Literature Review and Classification,” IFAC-PapersOnLine, vol. 51, no. 11, pp. 1016-1022, 2018.
   [CrossRef] [Google Scholar] [Publisher Link]
3. Chiara Cimino, Elisa Negri, and Luca Fumagalli, “Review of Digital Twin Applications in Manufacturing,” Computers in Industry, vol. 113, pp. 1-15, 2019.
   [CrossRef] [Google Scholar] [Publisher Link]
4. Yuqian Lu et al., “Digital Twin-Driven Smart Manufacturing: Connotation, Reference Model, Applications and Research Issues,” Robotics and Computer-Integrated Manufacturing, vol. 61, pp. 1-14, 2020.
   [CrossRef] [Google Scholar] [Publisher Link]
5. Radu F. Babiceanu, and Remzi Seker, “Cyber Resilience Protection for Industrial Internet of Things: A Software-Defined Networking Approach,” Computers in Industry, vol. 104, pp. 47-58, 2019.
   [CrossRef] [Google Scholar] [Publisher Link]
6. Diego Maldonado et al., “Multi-Agent Systems: A Survey about its Components, Framework and Workflow,” IEEE Access, vol. 12, pp. 80950-80975, 2024.
   [CrossRef] [Google Scholar] [Publisher Link]
7. Huaqing Liang, and Xunhe Yin, “Self-Healing Control: Review, Framework, and Prospect,” IEEE Access, vol. 11, pp. 79495-79512, 2023.
   [CrossRef] [Google Scholar] [Publisher Link]
8. Arunkumar Arulappan et al., “DQN Approach for Adaptive Self-Healing of VNFs in Cloud-Native Network,” IEEE Access, vol. 12, pp. 34489-34504, 2024.
   [CrossRef] [Google Scholar] [Publisher Link]
9. Aditya P. Mathur, and Nils Ole Tippenhauer, “SWaT: A Water Treatment Testbed for Research and Training on ICS Security,” 2026 International Workshop on Cyber-physical Systems for Smart Water Networks, Vienna, Austria, pp. 31-36, 2016.
   [CrossRef] [Google Scholar] [Publisher Link]
10. Nour Moustafa, “A New Distributed Architecture for Evaluating AI-based Security Systems at the Edge: Network TON_IoT Datasets,” Sustainable Cities and Society, vol. 72, pp. 1-36, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]
11. Mohamed Amine Ferrag et al., “Edge-IIoTset: A New Comprehensive Realistic Cyber Security Dataset of IoT and IIoT Applications for Centralized and Federated Learning,” IEEE Access, vol. 10, pp. 40281-40306, 2022.
    [CrossRef] [Google Scholar] [Publisher Link]
12. John Schulman et al., “Proximal Policy Optimization Algorithms,” arXiv preprint, pp. 1-12, 2017.
    [CrossRef] [Google Scholar] [Publisher Link]
13. Michael Wooldridge, An Introduction to MultiAgent Systems, 2^nd ed., John Wiley and Sons, 2009.
    [Google Scholar] [Publisher Link]
14. David Jones et al., “Characterising the Digital Twin: A Systematic Literature Review,” CIRP Journal of Manufacturing Science and Technology, vol. 29, pp. 36-52, 2020.
    [CrossRef] [Google Scholar] [Publisher Link]
15. Concetta Semeraro et al., “Digital Twin Paradigm: A Systematic Literature Review,” Computers in Industry, vol. 130, pp. 1-23, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]