Intellectual Property India Publishes Patent Application for Federated Tinyml-Based Privacy-Preserving Iot System For Distributed Anomaly Detection And Collaborative Intelligence.

MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641127010 A) filed by Padma Rajani; K. Akshitha; Dr. N. Srihari Rao; Dr. Narasimha Chary Ch; Dr. Naveen Reddy Nemili; Dr S. Srinivas; and Pillalamarri Laxman on October 25, 2026, for Federated Tinyml-Based Privacy-Preserving Iot System For Distributed Anomaly Detection And Collaborative Intelligence..

Inventors include Padma Rajani; K. Akshitha; Dr. N. Srihari Rao; Dr. Narasimha Chary Ch; Dr. Naveen Reddy Nemili; Dr S. Srinivas; and Pillalamarri Laxman.

The application for the patent was published on October 30, 2026, under issue no. 44/2026.

Abstract: ABSTRACT OF THE INVENTION The present invention discloses a federated TinyML-based privacy-preserving IoT system for distributed anomaly detection and collaborative intelligence. The system integrates lightweight machine learning models optimized for resource-constrained microcontrollers with a federated learning framework that enables multiple IoT devices to collaboratively improve a shared anomaly detection model without exchanging raw sensor data. Each edge device performs local inference and training using TinyML techniques such as model quantization and pruning, ensuring real-time anomaly detection with minimal energy and memory overhead. Model updates are protected by privacy-enhancing mechanisms including differential privacy and optional secure aggregation before being sent to a central or hierarchical coordinator. The coordinator aggregates the updates to produce an improved global model that is redistributed to participating devices. A collaborative intelligence component further allows the sharing of high-level anomaly insights, enabling the network to adapt to novel patterns observed in only a subset of devices. An adaptive scheduler optimizes communication rounds based on anomaly prevalence, network status, and residual energy. The invention offers significant advantages in privacy, latency, scalability, and suitability for battery-powered IoT deployments across industrial, healthcare, and smart infrastructure applications. By keeping sensitive data local while continuously enhancing detection performance through secure collaboration, the system provides a practical solution to the challenges of centralized IoT analytics.

Disclaimer: Curated by HT Syndication.