MUMBAI, India, July 30 -- Intellectual Property India has published a patent application (202641088010 A) filed by Dr. S. Muthukumar; Divesh Singh; Dr. Rakesh Sharma; Dr Badarla Anil; Rishoona R; Subashree V; Madhava Rao Karri; and Dr. A. Josephin Arockia Dhivya on July 18, 2026, for Edge Ai-Based Self-Healing Intrusion Detection And Response System For Industrial Iot Networks.
Inventors include Dr. S. Muthukumar; Divesh Singh; Dr. Rakesh Sharma; Dr Badarla Anil; Rishoona R; Subashree V; Madhava Rao Karri; and Dr. A. Josephin Arockia Dhivya.
The application for the patent was published on July 24, 2026, under issue no. 30/2026.
Abstract: ABSTRACT Edge AI-Based Self-Healing Intrusion Detection and Response System for Industrial IoT Networks The present invention discloses an Edge AI-Based Self-Healing Intrusion Detection and Response System for Industrial IoT (IIoT) Networks that enhances cybersecurity, resilience, and operational continuity in smart manufacturing, critical infrastructure, and industrial automation environments. The proposed system integrates distributed edge computing, artificial intelligence, federated learning, and autonomous recovery mechanisms to detect, classify, mitigate, and recover from cyber threats in real time without excessive dependence on centralized cloud infrastructure. Industrial edge nodes continuously monitor network traffic, device behavior, communication protocols, and process parameters using lightweight deep learning and anomaly detection models optimized for resource-constrained environments. The system identifies malicious activities such as malware propagation, ransomware attacks, unauthorized access, denial-of-service attacks, data manipulation, insider threats, and zero-day exploits by combining behavioral analytics, signature-based verification, and adaptive machine learning techniques. Upon threat detection, the self-healing engine autonomously isolates compromised devices, reconfigures communication routes, updates security policies, restores trusted configurations, and dynamically reallocates workloads to healthy nodes to maintain uninterrupted industrial operations. A secure federated learning framework enables collaborative intelligence among distributed edge devices while preserving data privacy and minimizing communication overhead. Blockchain-assisted integrity verification further secures event logs and response actions against tampering. The system continuously learns from evolving attack patterns through reinforcement learning, thereby improving detection accuracy and reducing false alarms over time. The proposed invention supports heterogeneous Industrial IoT devices, programmable logic controllers, sensors, actuators, and industrial communication protocols, ensuring seamless deployment across legacy and modern industrial environments. By combining edge intelligence, autonomous cyber defense, predictive analytics, and self-healing network orchestration, the invention significantly enhances cybersecurity resilience, minimizes downtime, protects critical industrial assets, and ensures reliable, secure, and continuous operation of Industry 4.0 ecosystems against sophisticated and rapidly evolving cyber threats.
Disclaimer: Curated by HT Syndication.
