MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641123004 A) filed by Sathyabama Institute Of Science And Technology on October 14, 2026, for Federated Learning-Based Privacy-Preserving Embedded Iot Network For Smart Infrastructure.
Inventors include Dr. D. Sheema; Dr. S. Bangaru Kamatchi; Dr. M. P. Vaishnnave; Dr. Aiswarya S; Dr. K. Madhushri; Dr. J. Jeslin Shanthamalar; and Dr. Lavanya P. M.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: The present invention discloses a federated learning-based, privacy-preserving embedded Internet of Things network for intelligent monitoring and management of smart infrastructure. The system comprises distributed sensor-enabled embedded nodes that collect infrastructure data and locally train machine learning models without transmitting raw data to a central server. Privacy-protected model updates are securely communicated to a federated aggregation unit to generate an improved global model. The system incorporates secure aggregation, differential privacy, authentication, anomaly detection, and adaptive communication mechanisms to enhance data security, reduce bandwidth usage, and improve computational efficiency. The proposed framework enables real-time detection of structural faults, equipment failures, environmental hazards, energy inefficiencies, and cyber threats. It provides a scalable, secure, and energy-efficient solution for smart buildings, transportation networks, industrial facilities, power systems, and urban infrastructure.
Disclaimer: Curated by HT Syndication.
